CINXE.COM

<!doctype html><html lang="en-us"> <head> <meta name="viewport" content="width=device-width,initial-scale=1,maximum-scale=5"> <meta charset="utf-8"> <meta http-equiv="Accept-CH" content="DPR"> <link rel="preconnect" href="https://static.mongodb.com" crossorigin /> <link rel="dns-prefetch" href="https://static.mongodb.com" /> <link rel="preconnect" href="https://webassets.mongodb.com" /> <link rel="dns-prefetch" href="https://webassets.mongodb.com" /> <link rel="preconnect" href="https://webimages.mongodb.com" /> <link rel="dns-prefetch" href="https://webimages.mongodb.com" /> <link rel="preconnect" href="https://cdn.cookielaw.org" /> <link rel="dns-prefetch" href="https://cdn.cookielaw.org" /> <link rel="preload" href="https://static.mongodb.com/com/fonts/DINWeb-Bold.woff" as="font" type="font/woff" crossorigin /> <link rel="preload" href="https://static.mongodb.com/com/fonts/EuclidCircularA-Regular-WebXL.woff2" as="font" type="font/woff2" crossorigin /> <link rel="preload" href="https://static.mongodb.com/com/fonts/EuclidCircularA-Medium-WebXL.woff2" as="font" type="font/woff2" crossorigin /> <title>What Are Vector Embeddings? | MongoDB | MongoDB</title> <meta property="og:type" content="article"> <meta property="og:site_name" content="MongoDB"> <meta property="og:title" content="What Are Vector Embeddings? | MongoDB"> <meta property="og:url" content="https://www.mongodb.com/resources/basics/vector-embeddings"> <meta property="og:image" content="http://s3.amazonaws.com/info-mongodb-com/_com_assets/cms/kuzt9r42or1fxvlq2-Meta_Generic.png"> <meta property="og:image:secure_url" content="https://webimages.mongodb.com/_com_assets/cms/kuzt9r42or1fxvlq2-Meta_Generic.png"> <meta name="description" content="Learn the basics of vector embeddings, its role in AI, and how MongoDB utilizes this technology."> <meta property="og:description" content="Learn the basics of vector embeddings, its role in AI, and how MongoDB utilizes this technology."> <meta name="segment-site-verification" content="H9hNimEbN3E66ZW2Xe50qbKSSivU8oDk"> <link rel="canonical" href="https://www.mongodb.com/resources/basics/vector-embeddings"> <meta name="twitter:card" content="summary_large_image"> <meta name="twitter:site" content="@mongodb"> <meta name="twitter:title" content="What Are Vector Embeddings? | MongoDB"> <meta name="twitter:description" content="Learn the basics of vector embeddings, its role in AI, and how MongoDB utilizes this technology."> <link rel="icon" href="/assets/images/global/favicon.ico" type="image/x-icon"> <link rel="shortcut icon" href="/assets/images/global/favicon.ico"> <link rel="stylesheet" href="https://static.mongodb.com/com/mongodb-general.3758a3cb0a171e8afa58c94042d87567.css"> <script type="application/ld+json"> {"@context":"http://schema.org","@type":"Organization","name":"MongoDB","url":"https://www.mongodb.com","logo":"https://webassets.mongodb.com/_com_assets/cms/mongodb_logo1-76twgcu2dm.png"} </script> <script> window.Intercom = function () { window.Intercom.c(arguments) } window.Intercom.q = [] window.Intercom.c = function (args) { window.Intercom.q.push(args) } </script> <link rel="preload" href="//cdn.optimizely.com/js/15508090763.js" as="script"> <link rel="preconnect" href="//logx.optimizely.com"> <script src="https://cdn.optimizely.com/js/15508090763.js" ></script> <script> // Can be removed once we stop supporting legacy edge. window.globalThis = window.globalThis || window </script> <script async>Number.isNaN = Number.isNaN || function (x) { return x !== x }</script> <script> !function(e,n){var t=document.createElement("script"),o=null,x="pathway";t.async=!0,t.src='https://'+x+'.mongodb.com/'+(e?x+'-debug.js':''),document.head.append(t),t.addEventListener("load",function(){o=window.pathway.default,(n&&o.configure(n)),o.createProfile("mongodbcom").load(),window.segment=o})}(); </script> <script async src="https://cdn.jsdelivr.net/npm/smoothscroll-polyfill@0.4.4/dist/smoothscroll.min.js"></script>  <script> window.SENTRY = { release: 'commit: 2e2d9a4', environment: 'production' } window.BUGSNAG = { apiKey: '85488288cc4942da2965a8b7e07bbf38', appVersion: 'commit: 2e2d9a4', releaseStage: 'production' } </script> <script> if (global === undefined) { var global = window } </script> <script type="module" crossorigin="anonymous" async src="https://static.mongodb.com/com/mdb-components.3811a9da.js"></script> <script type="module" crossorigin="anonymous" async src="https://static.mongodb.com/com/import-run.d5b02fd3.js"></script> <script type="module" crossorigin="anonymous" async src="https://static.mongodb.com/com/report-error.42f17b29.js"></script> <script type="module" crossorigin="anonymous" async src="https://static.mongodb.com/com/bootstrap-editor.5019f911.js"></script> <style data-styled="fJwnBQ obHHn emWlgh" data-styled-version="4.4.1"> /* sc-component-id: Article__ArticleStyles-sc-14w82yl-0 */ .obHHn pre > code{background:black;overflow-x:auto;} /* sc-component-id: Details__Container-sc-wfooue-0 */ .emWlgh{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;border-radius:4px;cursor:pointer;box-shadow:0 2px 7px 0 rgba(0,0,0,0.2);width:100%;background-color:white;margin:20px 0;padding-left:30px;} /* sc-component-id: editor-page__ContentWrapper-sc-1rxkxf7-0 */ @media (min-width:500px){.fJwnBQ{min-height:calc(100vh - 70px);}}</style> </head> <body> <noscript> <iframe src="https://obseu.michiganrobotflower.com/ns/1026a1528f8727653fd96984e7b20 597.html?ch=cheq4ppc" width="0" height="0" style="display:none"></iframe> </noscript>  <noscript> <iframe src="https://www.googletagmanager.com/ns.html?id=GTM-GDFN&nojscript=true" style="display:none"></iframe> </noscript>  <div class="react-root"><div><div class="editor-page__ContentWrapper-sc-1rxkxf7-0 fJwnBQ"><div class="pencil-banner-no-underline"><style data-emotion="css qdug2f">.css-qdug2f{position:relative;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;min-height:40px;height:auto;box-sizing:border-box;overflow:hidden;width:100%;cursor:pointer;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;padding:8px 24px;-webkit-text-decoration:none;text-decoration:none;color:#00684A;background-color:#00684A;}@media screen and (min-width: 460px){.css-qdug2f{height:40px;}}@media screen and (min-width: 1024px){.css-qdug2f{padding-left:48px;padding-right:48px;}}.css-qdug2f:hover mark{-webkit-text-decoration:underline;text-decoration:underline;}</style><div class="css-qdug2f"><style data-emotion="css bz3nwe">.css-bz3nwe{max-width:1420px;}</style><div class="css-bz3nwe"><style data-emotion="css j69nk4">.css-j69nk4{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;gap:gap-inc30;-webkit-text-decoration:none;text-decoration:none;max-width:100%;}.css-j69nk4:hover{-webkit-text-decoration:none;text-decoration:none;}.css-j69nk4:hover mark{-webkit-text-decoration:underline;text-decoration:underline;}@media screen and (min-width: 460px){.css-j69nk4{max-width:unset;}}</style><a href="https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner" tabIndex="0" class="css-j69nk4"><style data-emotion="css 4yi48v">.css-4yi48v{height:24px;line-height:24px;padding-left:12px;padding-right:12px;padding-top:0;padding-bottom:0;margin-right:16px;font-size:9px;white-space:nowrap;font-weight:600;}</style><style data-emotion="css crof2h">.css-crof2h{font-weight:600;text-transform:uppercase;font-family:Source Code Pro;line-height:16px;font-size:12px;display:inline-block;border-radius:999px;padding:4px 16px;letter-spacing:2.5px;color:#001E2B;background-color:#B1FF05;height:24px;line-height:24px;padding-left:12px;padding-right:12px;padding-top:0;padding-bottom:0;margin-right:16px;font-size:9px;white-space:nowrap;font-weight:600;}</style><span class="css-crof2h">Event</span><style data-emotion="css 36hkg1">.css-36hkg1{-webkit-text-decoration:none;text-decoration:none;color:white;font-weight:400;font-size:12px;line-height:18px;font-family:Source Code Pro,Noto Sans KR,Noto Sans SC,Noto Sans JP;overflow:hidden;text-overflow:ellipsis;}@media screen and (min-width: 768px){.css-36hkg1{max-width:85vw;}}.css-36hkg1 mark{color:#E9FF99;background-color:transparent;}.css-36hkg1>span{overflow:inherit;text-overflow:inherit;}@media (max-width: 767px){.css-36hkg1{font-weight:400;font-size:12px;line-height:16px;}.css-36hkg1>span{white-space:normal;-webkit-box-orient:vertical;-webkit-line-clamp:2;}.css-36hkg1>span:first-child{display:none;}.css-36hkg1>span:last-child{display:-webkit-box;}}@media (min-width: 768px){.css-36hkg1{font-weight:400;font-size:12px;line-height:16px;}.css-36hkg1>span{white-space:nowrap;}.css-36hkg1>span:first-child{display:block;}.css-36hkg1>span:last-child{display:none;}}</style><style data-emotion="css m8dkdf">.css-m8dkdf{margin:0;color:#001E2B;font-family:Euclid Circular A;font-size:14px;line-height:16px;-webkit-text-decoration:none;text-decoration:none;color:white;font-weight:400;font-size:12px;line-height:18px;font-family:Source Code Pro,Noto Sans KR,Noto Sans SC,Noto Sans JP;overflow:hidden;text-overflow:ellipsis;}@media screen and (min-width: 460px){.css-m8dkdf{font-size:14px;line-height:16px;}}@media screen and (min-width: 768px){.css-m8dkdf{font-size:14px;line-height:16px;}}@media screen and (min-width: 1024px){.css-m8dkdf{font-size:14px;line-height:16px;}}@media screen and (min-width: 768px){.css-m8dkdf{max-width:85vw;}}.css-m8dkdf mark{color:#E9FF99;background-color:transparent;}.css-m8dkdf>span{overflow:inherit;text-overflow:inherit;}@media (max-width: 767px){.css-m8dkdf{font-weight:400;font-size:12px;line-height:16px;}.css-m8dkdf>span{white-space:normal;-webkit-box-orient:vertical;-webkit-line-clamp:2;}.css-m8dkdf>span:first-child{display:none;}.css-m8dkdf>span:last-child{display:-webkit-box;}}@media (min-width: 768px){.css-m8dkdf{font-weight:400;font-size:12px;line-height:16px;}.css-m8dkdf>span{white-space:nowrap;}.css-m8dkdf>span:first-child{display:block;}.css-m8dkdf>span:last-child{display:none;}}</style><span class="css-m8dkdf"><span>Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >></mark></span><span>Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>>></mark></span></span></a></div></div></div><div style="position: sticky; top: 0px; z-index: 9999; width: 100%;"><style data-emotion="css-global wo4i12">@font-face{font-family:Akzidenz-Grotesk Std;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Regular-WebXL.woff2) format('woff2');font-weight:300;font-style:normal;font-display:swap;}@font-face{font-family:Akzidenz-Grotesk Std;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Medium-WebXL.woff2) format('woff2');font-weight:500;font-style:normal;font-display:swap;}</style><style data-emotion="css-global 17v57cw">@font-face{font-family:Euclid Circular A;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Regular-WebXL.woff2) format('woff2');font-weight:normal;font-display:swap;}@font-face{font-family:Euclid Circular A;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Medium-WebXL.woff2) format('woff2');font-weight:500;font-display:swap;}@font-face{font-family:MongoDB Value Serif;src:url(https://static.mongodb.com/com/fonts/MongoDBValueSerif-Regular.woff2) format('woff2');font-weight:normal;font-display:swap;}@font-face{font-family:MongoDB Value Serif;src:url(https://static.mongodb.com/com/fonts/MongoDBValueSerif-Medium.woff2) format('woff2');font-weight:500;font-display:swap;}@font-face{font-family:MongoDB Value Serif;src:url(https://static.mongodb.com/com/fonts/MongoDBValueSerif-Bold.woff2) format('woff2');font-weight:bold;font-display:swap;}@font-face{font-family:Source Code Pro;src:url(https://static.mongodb.com/com/fonts/SourceCodePro-Regular.ttf) format('truetype');font-weight:normal;font-display:swap;}@font-face{font-family:Source Code Pro;src:url(https://static.mongodb.com/com/fonts/SourceCodePro-Medium.ttf) format('truetype');font-weight:500;font-display:swap;}</style><style data-emotion="css 1ek23uy">.css-1ek23uy{width:100%;position:relative;top:0;left:0;z-index:1;}</style><nav role="navigation" class="css-1ek23uy"><style data-emotion="css hyy04k">.css-hyy04k{display:none;}@media screen and (min-width: 1024px){.css-hyy04k{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}}</style><style data-emotion="css x1631f">.css-x1631f{width:100%;background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;border-bottom:0;}@media screen and (min-width: 1024px){.css-x1631f{border-bottom:1px solid #b8c4c2;}}</style><style data-emotion="css 4eenyd">.css-4eenyd{box-sizing:border-box;margin:0;min-width:0;width:100%;background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;border-bottom:0;}@media screen and (min-width: 1024px){.css-4eenyd{border-bottom:1px solid #b8c4c2;}}</style><div class="css-4eenyd"><style data-emotion="css z9tlrl">.css-z9tlrl{width:100%;max-width:1512px;background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;font-weight:300;height:95px;padding-left:48px;padding-right:48px;display:none;}@media screen and (min-width: 1024px){.css-z9tlrl{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}}</style><style data-emotion="css frivp1">.css-frivp1{box-sizing:border-box;margin:0;min-width:0;width:100%;max-width:1512px;background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;font-weight:300;height:95px;padding-left:48px;padding-right:48px;display:none;}@media screen and (min-width: 1024px){.css-frivp1{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}}</style><div class="css-frivp1"><style data-emotion="css 15nzs5q">.css-15nzs5q{font-size:14px;line-height:37px;height:32px;width:126px;max-width:none;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}</style><a href="https://www.mongodb.com" class="css-15nzs5q"><style data-emotion="css 1qo9kov">.css-1qo9kov{width:384px;min-width:100px;font-size:14px;line-height:37px;height:32px;width:126px;max-width:none;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}</style><img src="https://webimages.mongodb.com/_com_assets/cms/kuyjf3vea2hg34taa-horizontal_default_slate_blue.svg?auto=format%252Ccompress" alt="MongoDB logo" width="126" height="32" class="css-1qo9kov" /></a><style data-emotion="css dc90up">.css-dc90up{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;width:100%;}</style><style data-emotion="css 1ppmow7">.css-1ppmow7{box-sizing:border-box;margin:0;min-width:0;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;width:100%;}</style><div class="header-desktop-buttons css-1ppmow7"><style data-emotion="css oc61gb">.css-oc61gb{background-color:#ffffff;opacity:0;-webkit-transition:opacity 250ms;transition:opacity 250ms;width:100%;height:95px;position:absolute;left:0;z-index:-1;top:0;}</style><style data-emotion="css 11yhye9">.css-11yhye9{box-sizing:border-box;margin:0;min-width:0;background-color:#ffffff;opacity:0;-webkit-transition:opacity 250ms;transition:opacity 250ms;width:100%;height:95px;position:absolute;left:0;z-index:-1;top:0;}</style><div class="css-11yhye9"></div><style data-emotion="css e3nr25">.css-e3nr25{position:absolute;visibility:hidden;z-index:-1;}</style><style data-emotion="css xapp63">.css-xapp63{box-sizing:border-box;margin:0;min-width:0;position:absolute;visibility:hidden;z-index:-1;}</style><div class="css-xapp63"><style data-emotion="css 1c69emu">.css-1c69emu{position:relative;width:90%;top:0;height:95px;display:none;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;padding-left:16px;padding-right:24px;}@media screen and (min-width: 1024px){.css-1c69emu{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}}</style><form role="search" method="GET" action="https://www.mongodb.com/search" class="css-1c69emu"><style data-emotion="css 1vufwc5">.css-1vufwc5{z-index:2;margin-top:-500px;-webkit-transition:margin-top 250ms;transition:margin-top 250ms;width:100%;display:grid;grid-template-columns:3fr 1fr;grid-gap:8px;padding-right:24px;}</style><style data-emotion="css 87svlz">.css-87svlz{box-sizing:border-box;margin:0;min-width:0;z-index:2;margin-top:-500px;-webkit-transition:margin-top 250ms;transition:margin-top 250ms;width:100%;display:grid;grid-template-columns:3fr 1fr;grid-gap:8px;padding-right:24px;}</style><div class="css-87svlz"><style data-emotion="css 36i4c2">.css-36i4c2{display:inline-block;position:relative;width:100%;z-index:2;}</style><div class="css-36i4c2"><style data-emotion="css 9vd5ud">.css-9vd5ud{width:100%;}</style><style data-emotion="css etrcff">.css-etrcff{--input-padding:16px;--invalid-input-padding:48px;--border-width:1px;background-color:#ffffff;border-color:#b8c4c2;color:#21313c;width:100%;}.css-etrcff{cursor:default;outline:none;font-size:16px;font-family:Akzidenz-Grotesk Std;font-weight:300;line-height:16px;border-radius:4px;border-style:solid;border-width:1px;box-sizing:border-box;height:48px;padding-top:calc(var(--input-padding) - var(--border-width));padding-bottom:calc(var(--input-padding) - var(--border-width));padding-left:calc(var(--input-padding) - var(--border-width));padding-right:calc(var(--input-padding) - var(--border-width));}.css-etrcff::-webkit-input-placeholder{font-weight:300;color:#21313c;}.css-etrcff::-moz-placeholder{font-weight:300;color:#21313c;}.css-etrcff:-ms-input-placeholder{font-weight:300;color:#21313c;}.css-etrcff::placeholder{font-weight:300;color:#21313c;}</style><input type="text" placeholder="Search products, whitepapers, & more..." value class="css-etrcff" /></div><style data-emotion="css 13va512">.css-13va512{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;min-width:298px;z-index:2;}</style><style data-emotion="css v2nqhr">.css-v2nqhr{box-sizing:border-box;margin:0;min-width:0;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;min-width:298px;z-index:2;}</style><div class="css-v2nqhr"><style data-emotion="css aef77t">.css-aef77t{width:250px;display:inline-block;position:relative;}</style><div class="css-aef77t"><style data-emotion="css v1v2x1">.css-v1v2x1{position:relative;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;width:250px;height:48px;border:1px solid #b8c4c2;border-radius:4px;box-sizing:border-box;cursor:pointer;font-weight:300;background-color:#ffffff;padding-left:0;color:#3d4f58;border-color:#b8c4c2;}</style><button role="button" type="button" class="css-v1v2x1"><style data-emotion="css 6k4l2y">.css-6k4l2y{font-family:Akzidenz-Grotesk Std;font-size:16px;line-height:16px;color:#21313c;padding-left:16px;width:250px;white-space:nowrap;overflow:hidden;text-overflow:ellipsis;text-align:left;}</style><span data-testid="selected-value" class="css-6k4l2y">General Information</span><style data-emotion="css 109dpaz">.css-109dpaz{padding:0 16px;}</style><div class="css-109dpaz"><style data-emotion="css 1yzkxhp">.css-1yzkxhp{-webkit-transform:rotateZ(0.5deg);-moz-transform:rotateZ(0.5deg);-ms-transform:rotateZ(0.5deg);transform:rotateZ(0.5deg);-webkit-transition:-webkit-transform .15s ease-in-out;transition:transform .15s ease-in-out;z-index:0;}</style><svg data-testid="icon" width="16" height="9" viewBox="0 0 16 9" fill="none" xmlns="http://www.w3.org/2000/svg" class="css-1yzkxhp"><style data-emotion="css 1tlq8q9">.css-1tlq8q9{stroke:#3d4f58;}</style><path d="M1.06689 0.799988L8.00023 7.73332L14.9336 0.799988" stroke-linecap="round" stroke-linejoin="round" class="css-1tlq8q9"></path></svg></div></button><style data-emotion="css 9vmgd1">.css-9vmgd1{visibility:hidden;position:absolute;z-index:1000;display:none;width:100%;min-width:250px;padding:16px;line-height:16px;font-size:16px;color:#21313c;font-family:Akzidenz-Grotesk Std;background-color:#ffffff;border:1px solid #b8c4c2;border-radius:8px;box-sizing:border-box;box-shadow:0px 3px 9px rgba(0, 0, 0, 0.15);}</style><div class="css-9vmgd1"><style data-emotion="css ac9zo2">.css-ac9zo2{list-style-type:none;margin:0;padding:0;}</style><ul data-testid="options" role="listbox" class="css-ac9zo2"><style data-emotion="css 11dtrvq">.css-11dtrvq{cursor:pointer;padding:8px;}.css-11dtrvq:not(:last-child){margin-bottom:8px;}.css-11dtrvq:hover{border-radius:2px;background-color:#e7f2eb;color:#09804c;}</style><li role="option" tabIndex="0" class="css-11dtrvq">General Information</li><li role="option" tabIndex="0" class="css-11dtrvq">Documentation</li><li role="option" tabIndex="0" class="css-11dtrvq">Developer Articles & Topics</li><li role="option" tabIndex="0" class="css-11dtrvq">Community Forums</li><li role="option" tabIndex="0" class="css-11dtrvq">Blog</li><li role="option" tabIndex="0" class="css-11dtrvq">University</li></ul></div></div><input type="hidden" id="addsearch" name="addsearch" value /><style data-emotion="css 1myrko">.css-1myrko{display:inline-block;}.css-1myrko:hover>button,.css-1myrko:hover>a{border-radius:40px;}</style><span class="css-1myrko"><style data-emotion="css 13l1z36">.css-13l1z36{width:100%;padding-top:16px;padding-bottom:16px;padding-left:32px;padding-right:32px;font-family:Euclid Circular A;font-size:16px;font-weight:500;border-radius:4px;line-height:16px;border:solid;border-width:1px;-webkit-text-decoration:none;text-decoration:none;display:inline-block;gap:8px;-webkit-transition:border-radius .15s;transition:border-radius .15s;color:#ffffff;stroke:#ffffff;fill:#ffffff;background-color:#001E2B;border-style:solid;padding:14px 14px;margin-left:4px;}@media screen and (min-width: 768px){.css-13l1z36{width:unset;}}.css-13l1z36:hover{cursor:pointer;-webkit-text-decoration:none;text-decoration:none;-webkit-transition:.1s;transition:.1s;}.css-13l1z36:active{box-shadow:0px 0px 0px 3px rgba(242, 197, 238, 1);-webkit-transition:.1s;transition:.1s;}.css-13l1z36:disabled,.css-13l1z36disabled:hover{color:#5d6c74;stroke:#5d6c74;fill:#5d6c74;border-color:#21313c;border-width:1px;border-radius:4px;cursor:not-allowed;}</style><button type="submit" tabIndex="0" data-track="true" class=" css-13l1z36"><style data-emotion="css r9fohf">.css-r9fohf{max-width:unset;}</style><img alt="search icon" src="https://webimages.mongodb.com/_com_assets/cms/lyj1z1iiimsre0lsz-search_updated_white.svg?auto=format%252Ccompress" width="18" height="18" class="css-r9fohf" /></button></span></div></div></form></div><style data-emotion="css 29u6e6">.css-29u6e6{margin:0;margin-left:16px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;height:95px;list-style:none;padding:0;opacity:1;pointer-events:initial;-webkit-transition:opacity 250ms;transition:opacity 250ms;position:unset;-webkit-animation:fadeIn 0.5s forwards;animation:fadeIn 0.5s forwards;}@media screen and (max-width: 1416px){.css-29u6e6{margin-left:12px;}}@global{@-webkit-keyframes fadeIn{from{opacity:0;}to{opacity:1;}}@keyframes fadeIn{from{opacity:0;}to{opacity:1;}}@-webkit-keyframes fadeOut{from{opacity:1;}to{opacity:0;}}@keyframes fadeOut{from{opacity:1;}to{opacity:0;}}}</style><ul class="header-desktop-nav-list css-29u6e6"><style data-emotion="css 37iurc">.css-37iurc{padding-left:20px;padding-right:20px;}</style><li class="header-nav-menu-item css-37iurc"><style data-emotion="css 8w0qf3">.css-8w0qf3{font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;position:relative;height:95px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-text-decoration:none;text-decoration:none;letter-spacing:unset;min-width:calc(64px + 4px);-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;white-space:nowrap;}.css-8w0qf3 .nav-chevron{margin-left:2px;fill:#5d6c74;-webkit-transition:-webkit-transform 250ms,fill 200ms;transition:transform 250ms,fill 200ms;}.css-8w0qf3:hover{-webkit-text-decoration:none;text-decoration:none;}.css-8w0qf3:focus-visible{outline:-webkit-focus-ring-color auto 1px;}</style><style data-emotion="css jxj2lf">.css-jxj2lf{font-family:Euclid Circular A;font-weight:300;cursor:pointer;background:none;border:none;padding:0px;font-size:16px;line-height:32px;color:#21313c;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;position:relative;height:95px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-text-decoration:none;text-decoration:none;letter-spacing:unset;min-width:calc(64px + 4px);-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;white-space:nowrap;}.css-jxj2lf .nav-chevron{margin-left:2px;fill:#5d6c74;-webkit-transition:-webkit-transform 250ms,fill 200ms;transition:transform 250ms,fill 200ms;}.css-jxj2lf:hover{-webkit-text-decoration:none;text-decoration:none;}.css-jxj2lf:focus-visible{outline:-webkit-focus-ring-color auto 1px;}</style><button tabIndex="0" data-track="true" class="css-jxj2lf"><style data-emotion="css 1edz58y">.css-1edz58y{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-1edz58y .textlink-default-text-class{color:#001E2B;border-bottom:0;-webkit-transition:color 200ms,text-shadow 200ms;transition:color 200ms,text-shadow 200ms;}.css-1edz58y .textlink-default-text-class:hover{border-bottom:2px solid #061621;}@media screen and (max-width: 1416px){.css-1edz58y .textlink-default-text-class{font-size:15px;line-height:15px;}}.css-1edz58y .textlink-default-text-class:hover{border-bottom:0;color:#00684A;text-shadow:0 0 1px rgba(0, 104, 74, 0.5);}.css-1edz58y .textlink-default-text-class:hover .nav-chevron{fill:#00684A;}.css-1edz58y .textlink-arrow-class{color:#00AA57;}.css-1edz58y .textlink-link-icon-class{color:#21313c;}.css-1edz58y:hover .textlink-text-class{color:#00AA57;-webkit-animation:linear 1 alternate;-webkit-animation-name:color;-webkit-animation-duration:300ms;}@-webkit-keyframes color{0%{color:#061621;}100%{left:green50;}}.css-1edz58y:hover .textlink-arrow-class{left:0;-webkit-animation:linear 1 alternate;-webkit-animation-name:runLink;-webkit-animation-duration:300ms;}@-webkit-keyframes runTitle{0%{left:0;}33%{left:25px;}66%{left:-25px;}100%{left:0;}}@-webkit-keyframes runLink{0%{left:-100px;}100%{left:0;}}</style><span class="css-1edz58y"><style data-emotion="css aq3x7l">.css-aq3x7l{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-size:16px;line-height:32px;color:#21313c;}</style><span class="textlink-default-text-class css-aq3x7l">Products<svg width="16" height="16" viewBox="0 0 16 16" fill="none" xmlns="http://www.w3.org/2000/svg" class="nav-chevron"><path fill-rule="evenodd" clip-rule="evenodd" d="M4.18362 5.76804C4.29823 5.65778 4.45193 5.59753 4.61093 5.60053C4.76994 5.60353 4.92126 5.66953 5.03162 5.78404L7.99962 8.93444L10.9676 5.78404C11.0216 5.72457 11.0869 5.67653 11.1598 5.64277C11.2326 5.609 11.3115 5.59021 11.3918 5.5875C11.472 5.58479 11.552 5.59821 11.627 5.62698C11.7019 5.65575 11.7704 5.69927 11.8282 5.75497C11.8861 5.81066 11.9321 5.87741 11.9637 5.95124C11.9953 6.02507 12.0117 6.10449 12.012 6.18478C12.0123 6.26508 11.9965 6.34463 11.9656 6.41871C11.9346 6.49278 11.889 6.55989 11.8316 6.61604L8.43162 10.216C8.37565 10.2741 8.30855 10.3203 8.23432 10.3519C8.1601 10.3834 8.08028 10.3997 7.99962 10.3997C7.91897 10.3997 7.83915 10.3834 7.76492 10.3519C7.6907 10.3203 7.62359 10.2741 7.56762 10.216L4.16762 6.61604C4.05736 6.50144 3.99711 6.34774 4.00011 6.18873C4.00311 6.02972 4.06911 5.87841 4.18362 5.76804Z" fill="inherit"></path></svg></span></span></button><style data-emotion="css 1e4twiw animation-1w559i1">.css-1e4twiw{border-radius:16px;box-shadow:0px 3px 20px 0px rgba(0, 0, 0, 0.15);position:absolute;overflow-y:auto;overflow-x:hidden;scrollbar-width:thin;max-height:calc(100vh - 88px);top:100%;left:34px;margin-top:-8px;background:#ffffff;-webkit-animation:animation-1w559i1 0.2s ease-in-out forwards;animation:animation-1w559i1 0.2s ease-in-out forwards;visibility:hidden;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none;pointer-events:none;}@-webkit-keyframes animation-1w559i1{from{opacity:1;}to{opacity:0;}}@keyframes animation-1w559i1{from{opacity:1;}to{opacity:0;}}</style><div class="css-1e4twiw"><style data-emotion="css 9ph9zl">.css-9ph9zl{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;}</style><div class="css-9ph9zl"><style data-emotion="css 1t6t43">.css-1t6t43{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;}</style><div class="css-1t6t43"><style data-emotion="css x20kx8">.css-x20kx8{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;padding-left:32px;padding-right:16px;padding-top:32px;}</style><div class="css-x20kx8"><style data-emotion="css xddzfi">.css-xddzfi{width:344px;max-width:344px;margin-right:40px;}</style><div class="css-xddzfi"><style data-emotion="css 1sdjll7">.css-1sdjll7{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-flex:1;-ms-flex:1;flex:1;padding-bottom:10px;}</style><div class="css-1sdjll7"><style data-emotion="css 18955fu">.css-18955fu{font-size:12px;line-height:12px;min-height:12px;font-weight:600;color:#3d4f58;margin-bottom:14px;text-transform:uppercase;}</style><div class="css-18955fu">Platform</div><style data-emotion="css 5tnj2v">.css-5tnj2v{padding-left:8px;padding-right:8px;padding-top:10px;padding-bottom:10px;margin-left:-8px;display:-webkit-inline-box;display:-webkit-inline-flex;display:-ms-inline-flexbox;display:inline-flex;border-radius:8px;line-height:16px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;margin-bottom:4px;}@media screen and (min-width: 1024px){.css-5tnj2v{margin-bottom:12px;}}.css-5tnj2v .menu-title{font-size:14px;line-height:14px;}@media screen and (min-width: 1024px){.css-5tnj2v .menu-title{font-size:16px;line-height:16px;}}.css-5tnj2v:hover{background:#fafbfc;}</style><style data-emotion="css 19929cr">.css-19929cr{font-family:Euclid Circular A;font-weight:500;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#006CFA;padding-left:8px;padding-right:8px;padding-top:10px;padding-bottom:10px;margin-left:-8px;display:-webkit-inline-box;display:-webkit-inline-flex;display:-ms-inline-flexbox;display:inline-flex;border-radius:8px;line-height:16px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;margin-bottom:4px;}.css-19929cr:hover{-webkit-text-decoration:none;text-decoration:none;}@media screen and (min-width: 1024px){.css-19929cr{margin-bottom:12px;}}.css-19929cr .menu-title{font-size:14px;line-height:14px;}@media screen and (min-width: 1024px){.css-19929cr .menu-title{font-size:16px;line-height:16px;}}.css-19929cr:hover{background:#fafbfc;}</style><a tabIndex="0" href="https://www.mongodb.com/atlas" target="_self" data-track="true" class="css-19929cr"><style data-emotion="css 1gdkn91">.css-1gdkn91{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-1gdkn91 .textlink-default-text-class{color:#001E2B;line-height:16px;border-bottom:0;font-weight:500;-webkit-align-items:flex-start;-webkit-box-align:flex-start;-ms-flex-align:flex-start;align-items:flex-start;-webkit-text-decoration:none;text-decoration:none;}.css-1gdkn91 .textlink-default-text-class:hover{border-bottom:0;-webkit-text-decoration:none;text-decoration:none;}.css-1gdkn91 .textlink-arrow-class{color:#001E2B;line-height:32px;}.css-1gdkn91 .textlink-link-icon-class{color:#001E2B;line-height:32px;}</style><span class="css-1gdkn91"><style data-emotion="css pbhol6">.css-pbhol6{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-size:16px;line-height:32px;color:#006CFA;}.css-pbhol6:hover{-webkit-text-decoration:none;text-decoration:none;}</style><span class="textlink-default-text-class css-pbhol6"><style data-emotion="css 6orj5s">.css-6orj5s{width:32px;height:32px;margin-right:12px;}</style><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_product_family.svg" alt="atlas_product_family" class=" css-6orj5s" /><style data-emotion="css x4n4mc">.css-x4n4mc{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;}</style><span class="css-x4n4mc"><span class="menu-title">Atlas</span><style data-emotion="css mmbp4l">.css-mmbp4l{display:inline-block;font-size:12px;line-height:15px;font-weight:400;color:#5d6c74;margin-top:4px;}@media screen and (min-width: 1024px){.css-mmbp4l{font-size:14px;line-height:18px;}}</style><span class="css-mmbp4l">Build on a developer data platform</span></span></span></span></a></div><div class="css-1sdjll7"><div class="css-18955fu">Platform Services</div><a tabIndex="0" href="https://www.mongodb.com/products/platform/atlas-database" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_database.svg" alt="atlas_database" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Database</span><span class="css-mmbp4l">Deploy a multi-cloud database</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/products/platform/atlas-search" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_search.svg" alt="atlas_search" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Search</span><span class="css-mmbp4l">Deliver engaging search experiences</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/products/platform/atlas-vector-search" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/mdb_vector_search.svg" alt="mdb_vector_search" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Vector Search</span><span class="css-mmbp4l">Design intelligent apps with gen AI</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/products/platform/atlas-stream-processing" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_stream_processing.svg" alt="atlas_stream_processing" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Stream Processing</span><span class="css-mmbp4l">Unify data in motion and data at rest</span></span></span></span></a></div></div><style data-emotion="css fpou7b">.css-fpou7b{width:344px;max-width:344px;}</style><div class="css-fpou7b"><div class="css-1sdjll7"><div class="css-18955fu">Self Managed</div><a tabIndex="0" href="https://www.mongodb.com/products/self-managed/enterprise-advanced" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/enterprise_advanced_product family.svg" alt="enterprise_advanced_product family" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Enterprise Advanced</span><span class="css-mmbp4l">Run and manage MongoDB yourself</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/products/self-managed/community-edition" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/community_edition_product_family.svg" alt="community_edition_product_family" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Community Edition</span><span class="css-mmbp4l">Develop locally with MongoDB</span></span></span></span></a></div><div class="css-1sdjll7"><div class="css-18955fu">Tools</div><a tabIndex="0" href="https://www.mongodb.com/products/tools/compass" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/mdb_compass.svg" alt="mdb_compass" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Compass</span><span class="css-mmbp4l">Work with MongoDB data in a GUI</span></span></span></span></a><a tabIndex="0" href="https://cloud.mongodb.com/ecosystem/?filter=integration" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_integration.svg" alt="atlas_integration" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Integrations</span><span class="css-mmbp4l">Integrations with third-party services</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/products/tools/relational-migrator" target="_self" data-track="true" class="css-19929cr"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/mdb_migrator.svg" alt="mdb_migrator" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Relational Migrator</span><span class="css-mmbp4l">Migrate to MongoDB with confidence</span></span></span></span></a></div></div></div><style data-emotion="css 1aq7tsw">.css-1aq7tsw{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}.css-1aq7tsw>div+div{border-left:1px solid #e7eeec;}</style><div class="css-1aq7tsw"><div class="css-9vd5ud"><style data-emotion="css 86227v">.css-86227v{padding-left:32px;padding-right:16px;padding-top:16px;padding-bottom:16px;border-top:1px solid #e7eeec;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;width:100%;box-sizing:border-box;line-height:14px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;}.css-86227v>span{width:100%;}.css-86227v:hover{background:#fafbfc;}.css-86227v:hover svg{opacity:1;}.css-86227v .menu-title{display:inline-block;}.css-86227v .menu-description{display:inline-block;font-size:12px;font-weight:400;color:#5d6c74;line-height:15px;margin-top:4px;}.css-86227v svg{stroke:#006CFA;opacity:0;-webkit-transition:opacity 0.3s ease-out;transition:opacity 0.3s ease-out;}</style><style data-emotion="css 7ejzmr">.css-7ejzmr{font-family:Euclid Circular A;font-weight:500;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#006CFA;width:100%;padding-left:32px;padding-right:16px;padding-top:16px;padding-bottom:16px;border-top:1px solid #e7eeec;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;width:100%;box-sizing:border-box;line-height:14px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;}.css-7ejzmr:hover{-webkit-text-decoration:none;text-decoration:none;}.css-7ejzmr>span{width:100%;}.css-7ejzmr:hover{background:#fafbfc;}.css-7ejzmr:hover svg{opacity:1;}.css-7ejzmr .menu-title{display:inline-block;}.css-7ejzmr .menu-description{display:inline-block;font-size:12px;font-weight:400;color:#5d6c74;line-height:15px;margin-top:4px;}.css-7ejzmr svg{stroke:#006CFA;opacity:0;-webkit-transition:opacity 0.3s ease-out;transition:opacity 0.3s ease-out;}</style><a tabIndex="0" href=" https://www.mongodb.com/products" target="_self" data-track="true" class="css-7ejzmr"><style data-emotion="css 1f7scwv">.css-1f7scwv{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-1f7scwv .textlink-default-text-class{color:#21313c;line-height:14px;border-bottom:0;-webkit-text-decoration:none;text-decoration:none;font-weight:500;font-size:14px;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;width:100%;}.css-1f7scwv .textlink-default-text-class:hover{border-bottom:0;-webkit-text-decoration:none;text-decoration:none;}.css-1f7scwv .textlink-arrow-class{color:#001E2B;line-height:32px;}.css-1f7scwv .textlink-link-icon-class{color:#001E2B;line-height:32px;}</style><span class="css-1f7scwv"><span class="textlink-default-text-class css-pbhol6"><style data-emotion="css 10mejol">.css-10mejol{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}</style><span class="css-10mejol"><span class="css-x4n4mc"><span class="menu-title">View All Products</span><span class="menu-description">Explore our full developer suite</span></span></span><style data-emotion="css vvcvyi">.css-vvcvyi{width:16px;height:16px;stroke:#3d4f58;fill:none;stroke-width:1px;}</style><svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div><div class="css-9vd5ud"><a tabIndex="0" href="https://www.mongodb.com/products/updates/version-release" target="_self" data-track="true" class="css-7ejzmr"><span class="css-1f7scwv"><span class="textlink-default-text-class css-pbhol6"><span class="css-10mejol"><span class="css-x4n4mc"><span class="menu-title">MongoDB 8.0</span><span class="menu-description">Our fastest version ever</span></span></span><svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div></div><style data-emotion="css 1p2ltr0">.css-1p2ltr0{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;border-left:6px solid #f5f7fA;}</style><div class="css-1p2ltr0"><style data-emotion="css 15n20pz">.css-15n20pz{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;min-width:208px;max-width:208px;}.css-15n20pz>.helper-section-item{border-bottom:6px solid #f5f7fA;border-top:0px;}.css-15n20pz>.helper-section-item:last-of-type{border-bottom:0;padding-bottom:24px;}</style><div class="css-15n20pz"><style data-emotion="css lkbdt0">.css-lkbdt0{width:100%;box-sizing:border-box;-webkit-flex:1;-ms-flex:1;flex:1;padding:24px;padding-bottom:18px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;}</style><div class="helper-section-item css-lkbdt0"><style data-emotion="css 1lxjpys">.css-1lxjpys{font-size:12px;font-weight:500;line-height:12px;color:#21313c;}</style><div class="css-1lxjpys">Build with MongoDB Atlas</div><style data-emotion="css 1qkz7n9">.css-1qkz7n9{font-weight:400;font-size:12px;margin-top:8px;color:#5d6c74;line-height:18px;}</style><div class="css-1qkz7n9">Get started for free in minutes</div><style data-emotion="css 7ysqtr">.css-7ysqtr{margin-top:16px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}</style><div class="css-7ysqtr"><style data-emotion="css 1u3h8p4">.css-1u3h8p4{font-size:12px;line-height:12px;width:100%;border:1px solid #b8c4c2;padding-top:7px;padding-bottom:7px;border-radius:999px;-webkit-transition:background 0.2s ease-in,border 0.2s ease-in;transition:background 0.2s ease-in,border 0.2s ease-in;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;color:#006CFA;}.css-1u3h8p4:hover{border-color:#006CFA;color:#ffffff;background:#006CFA;}</style><style data-emotion="css zh2ocw">.css-zh2ocw{font-family:Euclid Circular A;font-weight:500;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#006CFA;font-size:12px;line-height:12px;width:100%;border:1px solid #b8c4c2;padding-top:7px;padding-bottom:7px;border-radius:999px;-webkit-transition:background 0.2s ease-in,border 0.2s ease-in;transition:background 0.2s ease-in,border 0.2s ease-in;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;color:#006CFA;}.css-zh2ocw:hover{-webkit-text-decoration:none;text-decoration:none;}.css-zh2ocw:hover{border-color:#006CFA;color:#ffffff;background:#006CFA;}</style><a tabIndex="0" href="https://www.mongodb.com/cloud/atlas/register" target="_self" data-track="true" class="css-zh2ocw"><style data-emotion="css g5pq55">.css-g5pq55{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-g5pq55 .textlink-default-text-class{color:inherit;line-height:16px;border-bottom:0;font-size:12px;font-weight:500;-webkit-transition:color 0.2s ease-in;transition:color 0.2s ease-in;}.css-g5pq55 .textlink-default-text-class:hover{border-bottom:0;}.css-g5pq55 .textlink-arrow-class{color:#001E2B;line-height:32px;}.css-g5pq55 .textlink-link-icon-class{color:#001E2B;line-height:32px;}</style><span class="css-g5pq55"><span class="textlink-default-text-class css-pbhol6">Sign Up</span></span></a></div></div><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Test Enterprise Advanced</div><div class="css-1qkz7n9">Develop with MongoDB on-premises</div><div class="css-7ysqtr"><a tabIndex="0" href="https://www.mongodb.com/try/download/enterprise" target="_self" data-track="true" class="css-zh2ocw"><span class="css-g5pq55"><span class="textlink-default-text-class css-pbhol6">Download</span></span></a></div></div><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Try Community Edition</div><div class="css-1qkz7n9">Explore the latest version of MongoDB</div><div class="css-7ysqtr"><a tabIndex="0" href="https://www.mongodb.com/try/download/community" target="_self" data-track="true" class="css-zh2ocw"><span class="css-g5pq55"><span class="textlink-default-text-class css-pbhol6">Download</span></span></a></div></div></div></div></div></div></li><li class="header-nav-menu-item css-37iurc"><button tabIndex="0" data-track="true" class="css-jxj2lf"><span class="css-1edz58y"><span class="textlink-default-text-class css-aq3x7l">Resources<svg width="16" height="16" viewBox="0 0 16 16" fill="none" xmlns="http://www.w3.org/2000/svg" class="nav-chevron"><path fill-rule="evenodd" clip-rule="evenodd" d="M4.18362 5.76804C4.29823 5.65778 4.45193 5.59753 4.61093 5.60053C4.76994 5.60353 4.92126 5.66953 5.03162 5.78404L7.99962 8.93444L10.9676 5.78404C11.0216 5.72457 11.0869 5.67653 11.1598 5.64277C11.2326 5.609 11.3115 5.59021 11.3918 5.5875C11.472 5.58479 11.552 5.59821 11.627 5.62698C11.7019 5.65575 11.7704 5.69927 11.8282 5.75497C11.8861 5.81066 11.9321 5.87741 11.9637 5.95124C11.9953 6.02507 12.0117 6.10449 12.012 6.18478C12.0123 6.26508 11.9965 6.34463 11.9656 6.41871C11.9346 6.49278 11.889 6.55989 11.8316 6.61604L8.43162 10.216C8.37565 10.2741 8.30855 10.3203 8.23432 10.3519C8.1601 10.3834 8.08028 10.3997 7.99962 10.3997C7.91897 10.3997 7.83915 10.3834 7.76492 10.3519C7.6907 10.3203 7.62359 10.2741 7.56762 10.216L4.16762 6.61604C4.05736 6.50144 3.99711 6.34774 4.00011 6.18873C4.00311 6.02972 4.06911 5.87841 4.18362 5.76804Z" fill="inherit"></path></svg></span></span></button><div class="css-1e4twiw"><div class="css-9ph9zl"><div class="css-1t6t43"><div class="css-x20kx8"><style data-emotion="css cc0pau">.css-cc0pau{width:216px;max-width:216px;margin-right:32px;}</style><div class="css-cc0pau"><div class="css-1sdjll7"><div class="css-18955fu">Documentation</div><a tabIndex="0" href="https://www.mongodb.com/docs/atlas/" target="_self" data-track="true" class="css-19929cr"><style data-emotion="css 38hmqx">.css-38hmqx{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-38hmqx .textlink-default-text-class{color:#001E2B;line-height:16px;border-bottom:0;font-weight:500;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-text-decoration:none;text-decoration:none;}.css-38hmqx .textlink-default-text-class:hover{border-bottom:0;-webkit-text-decoration:none;text-decoration:none;}.css-38hmqx .textlink-arrow-class{color:#001E2B;line-height:32px;}.css-38hmqx .textlink-link-icon-class{color:#001E2B;line-height:32px;}</style><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Atlas Documentation</span><span class="css-mmbp4l">Get started using Atlas</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/docs/manual/" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Server Documentation</span><span class="css-mmbp4l">Learn to use MongoDB</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/docs/guides/" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Start With Guides</span><span class="css-mmbp4l">Get step-by-step guidance for key tasks</span></span></span></span></a></div></div><style data-emotion="css eho906">.css-eho906{width:216px;max-width:216px;margin-right:10px;}</style><div class="css-eho906"><div class="css-1sdjll7"><div class="css-18955fu"> </div><a tabIndex="0" href="https://www.mongodb.com/docs/tools-and-connectors/" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Tools and Connectors</span><span class="css-mmbp4l">Learn how to connect to MongoDB</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/docs/drivers/" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">MongoDB Drivers</span><span class="css-mmbp4l">Use drivers and libraries for MongoDB</span></span></span></span></a></div></div></div><div class="css-1aq7tsw"><div class="css-9vd5ud"><a tabIndex="0" href="https://www.mongodb.com/resources" target="_self" data-track="true" class="css-7ejzmr"><span class="css-1f7scwv"><span class="textlink-default-text-class css-pbhol6"><span class="css-10mejol"><span class="css-x4n4mc"><span class="menu-title">Resources Hub</span><span class="menu-description">Get help building the next big thing with MongoDB</span></span></span><svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div></div><div class="css-1p2ltr0"><style data-emotion="css 1a9krmi">.css-1a9krmi{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;min-width:208px;max-width:100%;padding-top:0px;padding-left:0px;padding-right:0px;box-sizing:border-box;}@media screen and (min-width: 1024px){.css-1a9krmi{max-width:320px;padding-top:32px;padding-left:24px;padding-right:24px;}}.css-1a9krmi>.helper-section-item{border-bottom:6px solid #f5f7fA;border-top:0px;}.css-1a9krmi>.helper-section-item:last-of-type{border-bottom:0;padding-bottom:24px;}</style><div class="css-1a9krmi"><div class="css-1sdjll7"><div class="css-18955fu">Connect</div><style data-emotion="css 1l423vo">.css-1l423vo{padding-left:8px;padding-right:8px;padding-top:10px;padding-bottom:10px;margin-left:-8px;display:-webkit-inline-box;display:-webkit-inline-flex;display:-ms-inline-flexbox;display:inline-flex;border-radius:8px;line-height:16px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;margin-bottom:4px;margin-right:-8px;}@media screen and (min-width: 1024px){.css-1l423vo{margin-bottom:12px;}}.css-1l423vo .menu-title{font-size:14px;line-height:14px;}@media screen and (min-width: 1024px){.css-1l423vo .menu-title{font-size:16px;line-height:16px;}}.css-1l423vo:hover{background:#fafbfc;}</style><style data-emotion="css qk955r">.css-qk955r{font-family:Euclid Circular A;font-weight:500;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#006CFA;padding-left:8px;padding-right:8px;padding-top:10px;padding-bottom:10px;margin-left:-8px;display:-webkit-inline-box;display:-webkit-inline-flex;display:-ms-inline-flexbox;display:inline-flex;border-radius:8px;line-height:16px;-webkit-transition:background 0.3s ease-out;transition:background 0.3s ease-out;margin-bottom:4px;margin-right:-8px;}.css-qk955r:hover{-webkit-text-decoration:none;text-decoration:none;}@media screen and (min-width: 1024px){.css-qk955r{margin-bottom:12px;}}.css-qk955r .menu-title{font-size:14px;line-height:14px;}@media screen and (min-width: 1024px){.css-qk955r .menu-title{font-size:16px;line-height:16px;}}.css-qk955r:hover{background:#fafbfc;}</style><a tabIndex="0" href="https://www.mongodb.com/developer/" target="_self" data-track="true" class="css-qk955r"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/atlas_product_family.svg" alt="atlas_product_family" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Developer Center</span><span class="css-mmbp4l">Explore a wide range of developer resources</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/community/" target="_self" data-track="true" class="css-qk955r"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/general_events_ask_the_experts.svg" alt="general_events_ask_the_experts" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Community</span><span class="css-mmbp4l">Join a global community of developers</span></span></span></span></a><a tabIndex="0" href="https://learn.mongodb.com/" target="_self" data-track="true" class="css-qk955r"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/general_content_tutorial.svg" alt="general_content_tutorial" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Courses and Certification</span><span class="css-mmbp4l">Learn for free from MongoDB</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/events" target="_self" data-track="true" class="css-qk955r"><span class="css-1gdkn91"><span class="textlink-default-text-class css-pbhol6"><img src="https://webimages.mongodb.com/_com_assets/icons/general_events_session.svg" alt="general_events_session" class=" css-6orj5s" /><span class="css-x4n4mc"><span class="menu-title">Events and Webinars</span><span class="css-mmbp4l">Find an event or webinar near you</span></span></span></span></a></div></div></div></div></div></li><li class="header-nav-menu-item css-37iurc"><button tabIndex="0" data-track="true" class="css-jxj2lf"><span class="css-1edz58y"><span class="textlink-default-text-class css-aq3x7l">Solutions<svg width="16" height="16" viewBox="0 0 16 16" fill="none" xmlns="http://www.w3.org/2000/svg" class="nav-chevron"><path fill-rule="evenodd" clip-rule="evenodd" d="M4.18362 5.76804C4.29823 5.65778 4.45193 5.59753 4.61093 5.60053C4.76994 5.60353 4.92126 5.66953 5.03162 5.78404L7.99962 8.93444L10.9676 5.78404C11.0216 5.72457 11.0869 5.67653 11.1598 5.64277C11.2326 5.609 11.3115 5.59021 11.3918 5.5875C11.472 5.58479 11.552 5.59821 11.627 5.62698C11.7019 5.65575 11.7704 5.69927 11.8282 5.75497C11.8861 5.81066 11.9321 5.87741 11.9637 5.95124C11.9953 6.02507 12.0117 6.10449 12.012 6.18478C12.0123 6.26508 11.9965 6.34463 11.9656 6.41871C11.9346 6.49278 11.889 6.55989 11.8316 6.61604L8.43162 10.216C8.37565 10.2741 8.30855 10.3203 8.23432 10.3519C8.1601 10.3834 8.08028 10.3997 7.99962 10.3997C7.91897 10.3997 7.83915 10.3834 7.76492 10.3519C7.6907 10.3203 7.62359 10.2741 7.56762 10.216L4.16762 6.61604C4.05736 6.50144 3.99711 6.34774 4.00011 6.18873C4.00311 6.02972 4.06911 5.87841 4.18362 5.76804Z" fill="inherit"></path></svg></span></span></button><div class="css-1e4twiw"><div class="css-9ph9zl"><div class="css-1t6t43"><div class="css-x20kx8"><style data-emotion="css 10ejslm">.css-10ejslm{width:192px;max-width:192px;margin-right:32px;}</style><div class="css-10ejslm"><div class="css-1sdjll7"><div class="css-18955fu">Use cases</div><a tabIndex="0" href="https://www.mongodb.com/solutions/use-cases/artificial-intelligence" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Artificial Intelligence</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/use-cases/payments" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Payments</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/use-cases/serverless" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Serverless Development</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/use-cases/gaming" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Gaming</span></span></span></span></a></div></div><style data-emotion="css 12h7cp9">.css-12h7cp9{width:192px;max-width:192px;margin-right:10px;}</style><div class="css-12h7cp9"><div class="css-1sdjll7"><div class="css-18955fu">Industries</div><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/financial-services" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Financial Services</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/telecommunications" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Telecommunications</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/healthcare" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Healthcare</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/retail" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Retail</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/public-sector" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Public Sector</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/solutions/industries/manufacturing" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Manufacturing</span></span></span></span></a></div></div></div><div class="css-1aq7tsw"><div class="css-9vd5ud"><a tabIndex="0" href="https://www.mongodb.com/solutions/solutions-library" target="_self" data-track="true" class="css-7ejzmr"><span class="css-1f7scwv"><span class="textlink-default-text-class css-pbhol6"><span class="css-10mejol"><span class="css-x4n4mc"><span class="menu-title">Solutions Library</span><span class="menu-description">Organized and tailored solutions to kick-start projects</span></span></span><svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div></div><div class="css-1p2ltr0"><div class="css-15n20pz"><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Developer Data Platform</div><style data-emotion="css jnux5f">.css-jnux5f{font-weight:400;font-size:12px;margin-top:8px;color:#5d6c74;line-height:15px;}</style><div class="css-jnux5f">Accelerate innovation at scale</div><div class="css-7ysqtr"><style data-emotion="css 1kx7zhg">.css-1kx7zhg{font-size:12px;line-height:12px;color:#006CFA;width:100%;}.css-1kx7zhg svg{-webkit-transition:-webkit-transform 0.2s;transition:transform 0.2s;}.css-1kx7zhg:hover svg{-webkit-transform:translateX(8px);-moz-transform:translateX(8px);-ms-transform:translateX(8px);transform:translateX(8px);}</style><style data-emotion="css d0mgft">.css-d0mgft{font-family:Euclid Circular A;font-weight:500;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#006CFA;font-size:12px;line-height:12px;color:#006CFA;width:100%;}.css-d0mgft:hover{-webkit-text-decoration:none;text-decoration:none;}.css-d0mgft svg{-webkit-transition:-webkit-transform 0.2s;transition:transform 0.2s;}.css-d0mgft:hover svg{-webkit-transform:translateX(8px);-moz-transform:translateX(8px);-ms-transform:translateX(8px);transform:translateX(8px);}</style><a tabIndex="0" href="https://www.mongodb.com/solutions/developer-data-platform" target="_self" data-track="true" class="css-d0mgft"><style data-emotion="css x0qvfd">.css-x0qvfd{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-x0qvfd .textlink-default-text-class{color:#006CFA;line-height:12px;border-bottom:0;font-size:12px;font-weight:500;width:auto;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;}.css-x0qvfd .textlink-default-text-class:hover{border-bottom:0;}@media screen and (min-width: 1024px){.css-x0qvfd .textlink-default-text-class{width:100%;}}.css-x0qvfd .textlink-default-text-class svg{stroke:#006CFA;margin-left:8px;}.css-x0qvfd .textlink-arrow-class{color:#001E2B;line-height:32px;}.css-x0qvfd .textlink-link-icon-class{color:#001E2B;line-height:32px;}</style><span class="css-x0qvfd"><span class="textlink-default-text-class css-pbhol6">Learn more<svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Startups and AI Innovators</div><div class="css-jnux5f">For world-changing ideas and AI pioneers</div><div class="css-7ysqtr"><a tabIndex="0" href="https://www.mongodb.com/solutions/startups" target="_self" data-track="true" class="css-d0mgft"><span class="css-x0qvfd"><span class="textlink-default-text-class css-pbhol6">Learn more<svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Customer Case Studies</div><div class="css-jnux5f">Hear directly from our users</div><div class="css-7ysqtr"><a tabIndex="0" href="https://www.mongodb.com/solutions/customer-case-studies" target="_self" data-track="true" class="css-d0mgft"><span class="css-x0qvfd"><span class="textlink-default-text-class css-pbhol6">See Stories<svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div></div></div></div></div></li><li class="header-nav-menu-item css-37iurc"><button tabIndex="0" data-track="true" class="css-jxj2lf"><span class="css-1edz58y"><span class="textlink-default-text-class css-aq3x7l">Company<svg width="16" height="16" viewBox="0 0 16 16" fill="none" xmlns="http://www.w3.org/2000/svg" class="nav-chevron"><path fill-rule="evenodd" clip-rule="evenodd" d="M4.18362 5.76804C4.29823 5.65778 4.45193 5.59753 4.61093 5.60053C4.76994 5.60353 4.92126 5.66953 5.03162 5.78404L7.99962 8.93444L10.9676 5.78404C11.0216 5.72457 11.0869 5.67653 11.1598 5.64277C11.2326 5.609 11.3115 5.59021 11.3918 5.5875C11.472 5.58479 11.552 5.59821 11.627 5.62698C11.7019 5.65575 11.7704 5.69927 11.8282 5.75497C11.8861 5.81066 11.9321 5.87741 11.9637 5.95124C11.9953 6.02507 12.0117 6.10449 12.012 6.18478C12.0123 6.26508 11.9965 6.34463 11.9656 6.41871C11.9346 6.49278 11.889 6.55989 11.8316 6.61604L8.43162 10.216C8.37565 10.2741 8.30855 10.3203 8.23432 10.3519C8.1601 10.3834 8.08028 10.3997 7.99962 10.3997C7.91897 10.3997 7.83915 10.3834 7.76492 10.3519C7.6907 10.3203 7.62359 10.2741 7.56762 10.216L4.16762 6.61604C4.05736 6.50144 3.99711 6.34774 4.00011 6.18873C4.00311 6.02972 4.06911 5.87841 4.18362 5.76804Z" fill="inherit"></path></svg></span></span></button><div class="css-1e4twiw"><div class="css-9ph9zl"><div class="css-1t6t43"><div class="css-x20kx8"><div class="css-cc0pau"><style data-emotion="css 1ydjkoa">.css-1ydjkoa{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-flex:1;-ms-flex:1;flex:1;padding-bottom:10px;margin-top:-10px;}</style><div class="css-1ydjkoa"><a tabIndex="0" href="https://www.mongodb.com/company/careers" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Careers</span><span class="css-mmbp4l">Start your next adventure</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/blog" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Blog</span><span class="css-mmbp4l">Read articles and announcements</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/company/newsroom" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Newsroom</span><span class="css-mmbp4l">Read press releases and news stories</span></span></span></span></a></div></div><div class="css-eho906"><div class="css-1ydjkoa"><a tabIndex="0" href="https://www.mongodb.com/company/partners" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Partners</span><span class="css-mmbp4l">Learn about our partner ecosystem</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/company/leadership" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Leadership</span><span class="css-mmbp4l">Meet our executive team</span></span></span></span></a><a tabIndex="0" href="https://www.mongodb.com/company" target="_self" data-track="true" class="css-19929cr"><span class="css-38hmqx"><span class="textlink-default-text-class css-pbhol6"><span class="css-x4n4mc"><span class="menu-title">Company</span><span class="css-mmbp4l">Learn more about who we are</span></span></span></span></a></div></div></div><div class="css-1aq7tsw"></div></div><div class="css-1p2ltr0"><div class="css-15n20pz"><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Contact Us</div><div class="css-jnux5f">Reach out to MongoDB</div><div class="css-7ysqtr"><a tabIndex="0" href="https://www.mongodb.com/company/contact" target="_self" data-track="true" class="css-d0mgft"><span class="css-x0qvfd"><span class="textlink-default-text-class css-pbhol6">Let’s chat<svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div><div class="helper-section-item css-lkbdt0"><div class="css-1lxjpys">Investors</div><div class="css-jnux5f">Visit our investor portal</div><div class="css-7ysqtr"><a tabIndex="0" href="https://investors.mongodb.com/" target="_self" data-track="true" class="css-d0mgft"><span class="css-x0qvfd"><span class="textlink-default-text-class css-pbhol6">Learn more<svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-vvcvyi"><title>arrow-right</title><path d="M17.3749 6.66663L26.6668 16M26.6668 16L17.3749 25.3333M26.6668 16H5.3335" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></span></span></a></div></div></div></div></div></div></li><li class="header-nav-menu-item css-37iurc"><style data-emotion="css 1h3lf6v">.css-1h3lf6v{font-family:Euclid Circular A;font-weight:300;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#21313c;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;position:relative;height:95px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-text-decoration:none;text-decoration:none;letter-spacing:unset;min-width:calc(64px + 4px);-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;white-space:nowrap;}.css-1h3lf6v:hover{-webkit-text-decoration:none;text-decoration:none;}.css-1h3lf6v .nav-chevron{margin-left:2px;fill:#5d6c74;-webkit-transition:-webkit-transform 250ms,fill 200ms;transition:transform 250ms,fill 200ms;}.css-1h3lf6v:hover{-webkit-text-decoration:none;text-decoration:none;}.css-1h3lf6v:focus-visible{outline:-webkit-focus-ring-color auto 1px;}</style><a tabIndex="0" href="https://www.mongodb.com/pricing" data-track="true" class="css-1h3lf6v"><span class="css-1edz58y"><style data-emotion="css nlb0hz">.css-nlb0hz{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-size:16px;line-height:32px;color:#21313c;}.css-nlb0hz:hover{-webkit-text-decoration:none;text-decoration:none;}</style><span class="textlink-default-text-class css-nlb0hz">Pricing</span></span></a></li></ul><style data-emotion="css 1pjb6cd">.css-1pjb6cd{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;-webkit-flex-shrink:0;-ms-flex-negative:0;flex-shrink:0;}</style><style data-emotion="css n8sm7x">.css-n8sm7x{box-sizing:border-box;margin:0;min-width:0;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;-webkit-flex-shrink:0;-ms-flex-negative:0;flex-shrink:0;}</style><div class="css-n8sm7x"><style data-emotion="css a59hv0">.css-a59hv0{border:none;background:none;outline:none;margin-right:32px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;z-index:1;height:95px;width:20px;}@media screen and (min-width: 1024px) and (max-width: 1280px){.css-a59hv0{width:24px;height:24px;}}.css-a59hv0>img{max-width:none;width:20px;height:20px;}@media screen and (min-width: 1024px) and (max-width: 1280px){.css-a59hv0>img{width:24px;height:24px;}}.css-a59hv0:focus-visible{outline:-webkit-focus-ring-color auto 1px;}</style><button aria-label="Open Search" class="header-desktop-button css-a59hv0"><img alt="Search" src="https://webimages.mongodb.com/_com_assets/cms/lyekm5ifrkqjod0wu-search_updated.svg?auto=format%252Ccompress" width="20" height="20" /></button><style data-emotion="css qcba1e">.css-qcba1e{letter-spacing:unset;margin-right:32px;-webkit-text-decoration:none;text-decoration:none;}@media screen and (max-width: 1280px){.css-qcba1e{display:none;}}</style><style data-emotion="css 1rolaoe">.css-1rolaoe{font-family:Euclid Circular A;font-weight:300;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#21313c;letter-spacing:unset;margin-right:32px;-webkit-text-decoration:none;text-decoration:none;}.css-1rolaoe:hover{-webkit-text-decoration:none;text-decoration:none;}@media screen and (max-width: 1280px){.css-1rolaoe{display:none;}}</style><a tabIndex="0" href="https://www.mongodb.com/services/support" data-track="true" class="header-desktop-link css-1rolaoe"><style data-emotion="css 1s6g5p9">.css-1s6g5p9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;text-align:left;}.css-1s6g5p9 .textlink-default-text-class{color:#001E2B;border-bottom:0;-webkit-transition:color 200ms,text-shadow 200ms;transition:color 200ms,text-shadow 200ms;}.css-1s6g5p9 .textlink-default-text-class:hover{color:#00684A;text-shadow:0 0 1px rgba(0, 104, 74, 0.5);border-bottom:0;}@media screen and (max-width: 1416px){.css-1s6g5p9 .textlink-default-text-class{font-size:15px;line-height:15px;}}.css-1s6g5p9 .textlink-arrow-class{color:#00AA57;}.css-1s6g5p9 .textlink-link-icon-class{color:#21313c;}.css-1s6g5p9:hover .textlink-text-class{color:#00AA57;-webkit-animation:linear 1 alternate;-webkit-animation-name:color;-webkit-animation-duration:300ms;}@-webkit-keyframes color{0%{color:#061621;}100%{left:green50;}}.css-1s6g5p9:hover .textlink-arrow-class{left:0;-webkit-animation:linear 1 alternate;-webkit-animation-name:runLink;-webkit-animation-duration:300ms;}@-webkit-keyframes runTitle{0%{left:0;}33%{left:25px;}66%{left:-25px;}100%{left:0;}}@-webkit-keyframes runLink{0%{left:-100px;}100%{left:0;}}</style><span class="css-1s6g5p9"><span class="textlink-default-text-class css-nlb0hz">Support</span></span></a><style data-emotion="css 1tokipu">.css-1tokipu{letter-spacing:unset;margin-right:40px;-webkit-text-decoration:none;text-decoration:none;}@media screen and (max-width: 1280px){.css-1tokipu{display:none;}}</style><style data-emotion="css 1l1k5">.css-1l1k5{font-family:Euclid Circular A;font-weight:300;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;display:inline-block;font-size:16px;line-height:32px;color:#21313c;letter-spacing:unset;margin-right:40px;-webkit-text-decoration:none;text-decoration:none;}.css-1l1k5:hover{-webkit-text-decoration:none;text-decoration:none;}@media screen and (max-width: 1280px){.css-1l1k5{display:none;}}</style><a tabIndex="0" href="https://account.mongodb.com/account/login" data-track="true" class="header-desktop-link css-1l1k5"><span class="css-1s6g5p9"><span class="textlink-default-text-class css-nlb0hz">Sign In</span></span></a><span class="css-1myrko"><style data-emotion="css vh2atz">.css-vh2atz{width:100%;padding-top:calc(16px - 1px);padding-bottom:calc(16px - 1px);padding-left:24px;padding-right:24px;font-family:Euclid Circular A;font-size:16px;font-weight:500;border-radius:4px;line-height:16px;border:solid;border-width:1px;-webkit-text-decoration:none;text-decoration:none;display:inline-block;gap:8px;-webkit-transition:border-radius .15s;transition:border-radius .15s;color:#001E2B;stroke:#001E2B;fill:#001E2B;border-color:#001E2B;background-color:#00ED64;margin-right:0;letter-spacing:0.16px;}@media screen and (min-width: 768px){.css-vh2atz{width:unset;}}.css-vh2atz:hover{cursor:pointer;-webkit-text-decoration:none;text-decoration:none;border-radius:40px;}.css-vh2atz:active,.css-vh2atz:focus{border-radius:999px;box-shadow:0px 0px 0px 3px rgba(242, 197, 238, 1);-webkit-transition:.1s;transition:.1s;}.css-vh2atz:disabled,.css-vh2atzdisabled:hover{background-color:#b8c4c2;cursor:not-allowed;color:#5d6c74;stroke:#5d6c74;fill:#5d6c74;box-shadow:0px 0px 0px 0px #000000;}@media screen and (max-width: 1280px){.css-vh2atz{display:none;}}</style><a tabIndex="0" href="https://www.mongodb.com/cloud/atlas/register" data-track="true" class="header-desktop-button css-vh2atz">Try Free</a></span><style data-emotion="css 1rurqt3">.css-1rurqt3{position:relative;}@media screen and (min-width: 1281px){.css-1rurqt3{display:none;}}</style><div class="css-1rurqt3"><style data-emotion="css 13aqjzy">.css-13aqjzy{display:inline-block;}</style><div class="css-13aqjzy"><style data-emotion="css 165p6md">.css-165p6md{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;border:solid;border-width:0;border-radius:50%;-webkit-transition:.15s;transition:.15s;background-color:transparent;stroke:#001E2B;fill:#001E2B;border-style:none;border-color:#00684A;width:36px;height:36px;cursor:pointer;}@media screen and (min-width: 1024px){.css-165p6md{border-style:solid;}}.css-165p6md:hover{cursor:pointer;stroke:#023430;fill:#023430;opacity:1;}@media screen and (min-width: 1024px){.css-165p6md:hover{background-color:#00684A;stroke:#ffffff;fill:#ffffff;opacity:1;}}.css-165p6md:active{box-shadow:0px 0px 0px 3px rgba(242, 197, 238, 1);-webkit-transition:.1s;transition:.1s;}.css-165p6md:disabled,.css-165p6mddisabled:hover{cursor:not-allowed;background-color:transparent;stroke:#00684A;fill:#00684A;box-shadow:0px 0px 0px 0px #000000;opacity:0.5;}.css-165p6md.active,.css-165p6md:hover{background:#e7eeec;fill:#001E2B;stroke:#001E2B;}.css-165p6md>svg{width:24px;height:24px;}</style><button tabIndex="0" data-track="true" class=" css-165p6md"><style data-emotion="css uqf5cc">.css-uqf5cc{width:16px;height:16px;stroke:inherit;fill:none;stroke-width:2px;}</style><svg viewBox="0 0 32 32" fill="none" xmlns="http://www.w3.org/2000/svg" class=" css-uqf5cc"><title>menu-vertical</title><path d="M17.3332 5.36936C17.3332 6.12564 16.7362 6.73872 15.9998 6.73872C15.2635 6.73872 14.6665 6.12564 14.6665 5.36936C14.6665 4.61308 15.2635 4 15.9998 4C16.7362 4 17.3332 4.61308 17.3332 5.36936Z" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path><path d="M17.3332 16C17.3332 16.7563 16.7362 17.3694 15.9998 17.3694C15.2635 17.3694 14.6665 16.7563 14.6665 16C14.6665 15.2437 15.2635 14.6306 15.9998 14.6306C16.7362 14.6306 17.3332 15.2437 17.3332 16Z" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path><path d="M17.3332 26.6306C17.3332 27.3869 16.7362 28 15.9998 28C15.2635 28 14.6665 27.3869 14.6665 26.6306C14.6665 25.8744 15.2635 25.2613 15.9998 25.2613C16.7362 25.2613 17.3332 25.8744 17.3332 26.6306Z" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></svg></button></div></div></div></div></div></div><style data-emotion="css vf2rex">.css-vf2rex{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}@media screen and (min-width: 1024px){.css-vf2rex{display:none;}}</style><style data-emotion="css 10o52y3">.css-10o52y3{background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;font-weight:300;overflow:hidden;height:56px;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;width:100%;z-index:999;position:relative;padding-left:24px;padding-right:24px;border-bottom:1px solid #b8c4c2;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}@media screen and (min-width: 1024px){.css-10o52y3{display:none;}}</style><style data-emotion="css 6easlo">.css-6easlo{box-sizing:border-box;margin:0;min-width:0;background-color:#ffffff;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;font-family:Euclid Circular A,Noto Sans KR,Noto Sans SC,Noto Sans JP;font-weight:300;overflow:hidden;height:56px;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;width:100%;z-index:999;position:relative;padding-left:24px;padding-right:24px;border-bottom:1px solid #b8c4c2;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;}@media screen and (min-width: 1024px){.css-6easlo{display:none;}}</style><div class="css-6easlo"><style data-emotion="css knbtqt">.css-knbtqt{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;font-size:14px;height:24px;width:95px;min-width:95px;max-width:none;}@media screen and (min-width: 768px){.css-knbtqt{min-width:126px;width:126px;height:32px;}}</style><a href="https://www.mongodb.com" class="css-knbtqt"><style data-emotion="css 3el0ca">.css-3el0ca{width:95px;min-width:100px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;font-size:14px;height:24px;width:95px;min-width:95px;max-width:none;}@media screen and (min-width: 768px){.css-3el0ca{min-width:126px;width:126px;height:32px;}}</style><img src="https://webimages.mongodb.com/_com_assets/cms/kuyjf3vea2hg34taa-horizontal_default_slate_blue.svg?auto=format%252Ccompress" alt="MongoDB logo" width="95px" height="24px" class="css-3el0ca" /></a><style data-emotion="css sk3y9d">.css-sk3y9d{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;position:relative;z-index:1;}</style><style data-emotion="css 85rf0r">.css-85rf0r{box-sizing:border-box;margin:0;min-width:0;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;position:relative;z-index:1;}</style><div class="css-85rf0r"><style data-emotion="css 1a7pihi">.css-1a7pihi{border:none;background:none;outline:none;margin-right:20px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;z-index:1;height:95px;width:16px;}.css-1a7pihi>img{width:16px;height:16px;max-width:none;}.css-1a7pihi:focus-visible{outline:-webkit-focus-ring-color auto 1px;}</style><button aria-label="Open Search" class="css-1a7pihi"><img alt="Search" src="https://webimages.mongodb.com/_com_assets/cms/lyekm5ifrkqjod0wu-search_updated.svg?auto=format%252Ccompress" width="16" height="16" /></button><style data-emotion="css 1mpxh5k">.css-1mpxh5k{border:none;background:none;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;cursor:pointer;height:56px;padding-left:2px;padding-right:2px;}</style><button aria-label="Open Links" class="css-1mpxh5k"><style data-emotion="css 1dd6xh2">.css-1dd6xh2>rect{-webkit-transition:all 0.3s ease-in-out,opacity 0.2s linear 0.1s;transition:all 0.3s ease-in-out,opacity 0.2s linear 0.1s;transform-origin:50% 50%;}.css-1dd6xh2>rect#top-line{-webkit-transform:translateY(-4.375px);-moz-transform:translateY(-4.375px);-ms-transform:translateY(-4.375px);transform:translateY(-4.375px);}.css-1dd6xh2>rect#bottom-line{-webkit-transform:translateY(4.375px);-moz-transform:translateY(4.375px);-ms-transform:translateY(4.375px);transform:translateY(4.375px);}.css-1dd6xh2.animating #top-line,.css-1dd6xh2.active #top-line{-webkit-transform:translateY(0);-moz-transform:translateY(0);-ms-transform:translateY(0);transform:translateY(0);}.css-1dd6xh2.animating #middle-line,.css-1dd6xh2.active #middle-line{opacity:0;}.css-1dd6xh2.animating #bottom-line,.css-1dd6xh2.active #bottom-line{-webkit-transform:translateY(0);-moz-transform:translateY(0);-ms-transform:translateY(0);transform:translateY(0);}.css-1dd6xh2.active:not(.animating) #top-line{-webkit-transform:translateY(0) rotate(-45deg);-moz-transform:translateY(0) rotate(-45deg);-ms-transform:translateY(0) rotate(-45deg);transform:translateY(0) rotate(-45deg);}.css-1dd6xh2.active:not(.animating) #middle-line{opacity:0;}.css-1dd6xh2.active:not(.animating) #bottom-line{-webkit-transform:translateY(0) rotate(-135deg);-moz-transform:translateY(0) rotate(-135deg);-ms-transform:translateY(0) rotate(-135deg);transform:translateY(0) rotate(-135deg);}</style><svg width="16" height="10" viewBox="0 0 16 10" xmlns="http://www.w3.org/2000/svg" overflow="visible" class=" css-1dd6xh2"><rect id="top-line" x="0.5" y="4.375" rx="0.625" ry="0.625" width="15" height="1.25" fill="#21313C"></rect><rect id="middle-line" x="0.5" y="4.375" rx="0.625" ry="0.625" width="15" height="1.25" fill="#21313C"></rect><rect id="bottom-line" x="0.5" y="4.375" rx="0.625" ry="0.625" width="15" height="1.25" fill="#21313C"></rect></svg></button></div></div></nav></div><div class="relative"><div class="absolute w-full h-full overflow-hidden" style="z-index: -1;"><div style="background-position: center; background-repeat: no-repeat; background-size: cover; z-index: -1;" class="lazyload absolute w-full h-full"></div></div><header class="relative overflow-hidden fl fl-center fl-col fl-wrap h-min-340"><div class="absolute w-full h-full"><div style="background-position: center; background-repeat: no-repeat; background-size: cover;" data-bg="https://webassets.mongodb.com/_com_assets/cms/header-au039oiqsr.svg" class="lazyload absolute w-full h-full"></div></div><div class="relative w-full w-max-770 p-20 txt-center"><h1 class="m-t-0 fnt-medium dark-green"><span>What are Vector Embeddings? </span></h1><div class="fl fl-center w-full"><a class="reset" target="_blank" href="https://www.mongodb.com/cloud/atlas/register"><button class="relative btn-green btn-sm m-10">Get started free</button></a></div></div></header></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><p>Vector embeddings are mathematical representations of text created by translating words or sentences into numbers — a language that computers can understand. They bridge the rich, nuanced world of human language (text, images, speeches, videos etc.) and the precise environment of machine learning models (numbers) by representing data points.</p><p><img alt="An image of vector embeddings including unstructured data and encoder." src="https://webimages.mongodb.com/_com_assets/cms/m3n5dd9oq8zrgpot7-vector-embedding.png?auto=format%252Ccompress" /></p><br /><p>Most often used in natural language processing (NLP), vector embeddings allow machine learning algorithms to analyze information much like humans but at a scale and speed far beyond our capabilities. Although they also work with images, audio processing, bioinformatics, and recommendation systems, this article will focus on word vector embeddings in natural language processing using a machine learning model. </p><br /><p><strong>Table of contents</strong></p><ul class="bullets"><li><a target="_self" href="#what-is-natural-language-processing">What is natural language processing?</a></li><li><a target="_self" href="#types-of-vector-embeddings-in-nlp">Types of vector embeddings in NLP</a></li><li><a target="_self" href="#what-is-a-vector-">What is a vector?</a></li><li><a target="_self" href="#vector-embeddings-in-multidimensional-spaces">Vector embeddings in multidimensional spaces</a></li><li><a target="_self" href="#dimensionality-in-vector-embeddings">Dimensionality in vector embeddings</a></li><li><a target="_self" href="#advantages-of-vector-embeddings-in-realworld-scenarios">Advantages of vector embeddings in real-world scenarios</a></li><li><a target="_self" href="#challenges-and-limitations">Challenges and limitations</a></li><li><a target="_self" href="#using-vector-embeddings-in-other-applications-">Using vector embeddings in other applications</a></li><li><a target="_self" href="#why-use-mongodb-atlas-vector-search-for-vector-similarity-search">Why use MongoDB Atlas Vector Search for vector similarity search?</a></li><li><a target="_self" href="#conclusion">Conclusion</a></li><li><a target="_self" href="#faqs">FAQs</a></li></ul></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="what-is-natural-language-processing" style="font-family: Euclid Circular A; font-weight: 500;">What is natural language processing?</h2><p>NLP is a type of artificial intelligence that uses vector embeddings in conjunction with machine learning algorithms to evaluate, understand, and interpret human language. This combination achieves comprehension and interaction that mirrors human ability but at a scale and speed far beyond our capabilities. NLP excels in tasks such as interpreting text from social media, translating languages, and powering conversational agents.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="types-of-vector-embeddings-in-nlp" style="font-family: Euclid Circular A; font-weight: 500;">Types of vector embeddings in NLP</h2><p>Below are a few examples of the diverse vector embedding techniques instrumental in advancing NLP, each bringing its strengths to various language understanding challenges.</p><ul class="bullets"><li><p><strong>Word2Vec</strong>: Developed by Google, Word2Vec captures the context of words within documents. It’s beneficial for tasks that require understanding word associations and meanings based on their usage in sentences.</p></li><li><p><strong>GloVe (global vectors for word representation)</strong>: GloVe is unique in its approach as it analyzes word co-occurrences over the whole corpus for training, enabling it to capture global statistics of words. It’s particularly useful for tasks that involve semantic similarity between words.</p></li><li><p><strong>BERT (Bidirectional Encoder Representations from Transformers)</strong>: Developed by Google, BERT represents a breakthrough in contextually aware embeddings. It looks at the context from both sides of a word in a sentence, making it highly effective for sophisticated tasks like sentiment analysis and question answering.</p></li></ul></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="what-is-a-vector-" style="font-family: Euclid Circular A; font-weight: 500;">What is a vector? </h2><p>To truly appreciate what vector embeddings are and how they work, it's essential first to understand what a vector is in this setting. Think of a vector as a point in space with direction and magnitude. It's like a dot on a map with specific coordinates. These data points aren't just random numbers; they represent different characteristics or features of the data types that the vector represents.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="vector-embeddings-in-multidimensional-spaces" style="font-family: Euclid Circular A; font-weight: 500;">Vector embeddings in multidimensional spaces</h2><p>Now that the basics of vectors and data points have been introduced, it's important to note that in the context of vector representations for text, more than one vector embedding or one dimension is used. When text — words, phrases, or entire documents — are converted into vectors, each piece of text plots as a point in a vast, multidimensional space. This space isn't like a typical three-dimensional space we're familiar with; it has many more dimensions, each representing a different aspect of the text's meaning or usage. </p><p>Imagine a map where words with similar meanings or usages are mapped closer together, making it easier to see their relationships or similar data points.</p><p>Turning text into vector embeddings is a game-changer for machine learning algorithms. These algorithms are great at dealing with numbers — they can spot patterns, make comparisons, and draw conclusions from numerical data.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="dimensionality-in-vector-embeddings" style="font-family: Euclid Circular A; font-weight: 500;">Dimensionality in vector embeddings</h2><p>Let's dig a little deeper into the concept of dimensionality, which we introduced above. Think of dimensionality in vector embeddings like the resolution of a photo. High-resolution photos are more detailed and precise, but they take up more space on your phone and require more processing power. Similarly, in vector embeddings, more dimensions mean that the representation of words or phrases can capture more details and nuances of language.</p><br /><h3 id="high-dimensional-embeddings">High-dimensional embeddings</h3><p>High-dimensional embeddings are like high-resolution photos. They have hundreds or even thousands of dimensions, allowing them to capture much information about a word or phrase. Each dimension can represent a different aspect of a word's meaning or use. This detailed representation is excellent for complex tasks in natural language processing, where understanding subtle differences in language is crucial. </p><p>However, like high-resolution photos, these embeddings require more computer memory and processing power. Also, there's a risk of "overfitting" — think of it like a camera that focuses on capturing every tiny detail and fails to recognize common, everyday objects. In machine learning, the model might get too tailored to its training data and perform poorly on new, unseen data.</p><br /><h3 id="low-dimensional-embeddings">Low-dimensional embeddings</h3><p>On the other hand, low-dimensional embeddings are like lower-resolution photos. They have fewer dimensions, so they use less computer memory and process more quickly, which is excellent for applications that need to run fast or have limited resources. But just like lower-resolution photos can miss finer details, these embeddings might not capture all the subtle nuances of language. Depending on the task, they provide a more general picture, which can sometimes be enough.</p><p>Choosing the proper dimensionality for creating vector embeddings is a balance. It's about weighing the need for detail against the need for efficiency and the ability of the model to perform well on new, unseen data. Finding the right balance often involves trial and error and depends on the specific task and the data. It's a crucial part of developing effective NLP solutions, requiring a thoughtful approach to meet both the linguistic needs of the task and the practical limitations of technology.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="advantages-of-vector-embeddings-in-realworld-scenarios" style="font-family: Euclid Circular A; font-weight: 500;">Advantages of vector embeddings in real-world scenarios</h2><p>Vector embeddings have opened up a world of possibilities in how machines interact with human language. They make technology more intuitive and natural, enriching interactions across digital platforms and tools. Below are a few applications highlighting how vector embeddings are used today. </p><br /><h3 id="sentiment-analysis">Sentiment analysis</h3><p>Sentiment analysis is like a digital mood ring. Businesses use it to understand how people feel about their products or services by analyzing the tone of customer reviews and social media posts. Vector embeddings help computers catch subtle emotional cues in text, distinguishing genuine praise from sarcasm, even when the words are similar.</p><br /><h3 id="machine-translation">Machine translation</h3><p>Vector embeddings are the backbone of translation apps. They help computers grasp the complexities and nuances of different languages. When a sentence is translated from one language to another, it's not just about swapping words; it's about conveying the same meaning, tone, and context. Vector embeddings are crucial in achieving this.</p><br /><h3 id="chatbots-and-virtual-assistants">Chatbots and virtual assistants</h3><p>Are you curious how virtual assistants like Siri or Alexa understand and respond to your queries so well? This functionality is primarily due to vector embeddings. They enable artificial intelligence (AI) systems to process what you're saying, figure out what you mean, and respond in a way that makes sense.</p><br /><h3 id="information-retrieval">Information retrieval</h3><p>This category covers everything from search engines to recommendation systems. Vector embeddings help these systems understand what is being searched, not just by matching keywords but by grasping the context of the query. This way, the information or recommendations will more likely be relevant.</p><br /><h3 id="text-classification">Text classification</h3><p>Text classification can filter emails, categorize news articles, and even tag social media posts. Vector embeddings assist in sorting text into different categories by understanding the underlying themes and topics, making it easier for algorithms to decide, for example, if an email is spam.</p><br /><h3 id="speech-recognition">Speech recognition</h3><p>Regarding converting spoken words into written text, vector embeddings play a crucial role. They help capture the spoken words accurately, considering how the same word can be pronounced or used in different contexts, leading to more accurate transcriptions.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="challenges-and-limitations" style="font-family: Euclid Circular A; font-weight: 500;">Challenges and limitations</h2><p>While vector embeddings are a powerful tool in NLP, they are not without their challenges. Addressing these issues is crucial for ensuring that these technologies are effective, fair, and up-to-date, requiring continuous effort and innovation in the field. Let's explore these challenges and limitations, especially in how vector embeddings interact with and process human language.</p><br /><h3 id="handling-out-of-vocabulary-words">Handling out-of-vocabulary words</h3><p>One of the trickiest issues with vector embeddings is dealing with words that the system has never seen before, often called "out-of-vocabulary" words. It's like encountering a word in a foreign language you've never learned. For a computer, these new words can be a fundamental stumbling block. The system might struggle to understand and place them correctly in the context of what it already knows.</p><br /><h3 id="bias-in-embeddings">Bias in embeddings</h3><p>Like humans, computers can be biased, especially when they learn from data reflecting human prejudices. When vector embeddings train on text data from the internet or other human-generated sources, they can inadvertently pick up and even amplify these biases. This possibility is significant because it can lead to unfair or stereotypical representations in various applications, like search engines or AI assistants.</p><br /><h3 id="complexity-of-maintaining-and-updating-models">Complexity of maintaining and updating models</h3><p>Keeping vector embedding models up-to-date and relevant is no small task. Language is constantly evolving — new words pop up, old words fade away, and meanings change. Ensuring these models stay current is like updating a constantly evolving dictionary. It requires ongoing work and resources, making it a complex and challenging aspect of working with vector embeddings.</p><br /><h3 id="contextual-ambiguity">Contextual ambiguity</h3><p>While vector embeddings are good at capturing meaning, they sometimes struggle with words with multiple meanings based on context. For instance, the word "bat" can refer to an animal or sports equipment, and without sufficient context, the model might not accurately capture the intended use.</p><br /><h3 id="resource-intensity">Resource intensity</h3><p>Training sophisticated vector embedding models requires significant computational resources, which can be a barrier, especially for smaller organizations or individual researchers who might not have access to the necessary computing power.</p><br /><h3 id="data-quality-and-availability">Data quality and availability</h3><p>The effectiveness of vector embeddings heavily depends on the quality and quantity of the training data. The embeddings might not be as accurate or helpful in languages or domains where data is scarce or of poor quality.</p><br /><h3 id="transferability-across-languages">Transferability across languages</h3><p>Vector embeddings trained in one language may not transfer well to another, especially for structurally different languages. These structural differences challenge multilingual applications or languages with limited resources.</p><br /><h3 id="model-interpretability">Model interpretability</h3><p>Understanding why a vector embedding model behaves a certain way or makes specific decisions can be challenging. This lack of interpretability can be a significant issue, especially in applications where understanding the model's reasoning is crucial.</p><br /><h3 id="scalability">Scalability</h3><p>As the amount of data and the complexity of tasks increase, scaling vector embedding models while maintaining performance and efficiency can be challenging.</p><br /><h3 id="dependency-on-training-data">Dependency on training data</h3><p>Vector embeddings can only be as good as their training data. If the training data is limited or biased, the embeddings will inherently reflect those limitations or biases.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="using-vector-embeddings-in-other-applications-" style="font-family: Euclid Circular A; font-weight: 500;">Using vector embeddings in other applications </h2><p>While vector embeddings are most prominently used in NLP, they are also used in other ways. Here are a few other areas where vector embeddings are employed. </p><br /><h3 id="computer-vision">Computer vision</h3><p>In image processing and computer vision, embeddings represent images or parts of images. Similar to how they capture the essence of words in NLP, embeddings in computer vision capture essential features of images, enabling tasks like image recognition, classification, and similarity detection.</p><br /><h3 id="recommendation-systems">Recommendation systems</h3><p>Vector embeddings also show up in recommendation systems, such as those on e-commerce or streaming platforms. They help understand user preferences and item characteristics by representing users and items in a vector space, enabling the system to make personalized recommendations based on similarity.</p><br /><h3 id="bioinformatics">Bioinformatics</h3><p>In bioinformatics, embeddings can represent biological data, such as gene sequences or protein structures. These embeddings help in various predictive tasks, like understanding gene function or protein-protein interactions.</p><br /><h3 id="graph-analysis">Graph analysis</h3><p>In network and graph analysis, embeddings represent nodes and edges of a graph, which is helpful in social network analysis, link prediction, and understanding the structure and dynamics of complex systems.</p><br /><h3 id="time-series-analysis">Time series analysis</h3><p>Vector embeddings work in analyzing time series data, such as financial market trends or sensor data, by capturing temporal patterns and dependencies in a vector space.</p><p>These diverse applications show that the concept of embeddings is a versatile tool in the broader field of machine learning and data science, not limited to just text and language processing.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="why-use-mongodb-atlas-vector-search-for-vector-similarity-search" style="font-family: Euclid Circular A; font-weight: 500;">Why use MongoDB Atlas Vector Search for vector similarity search?</h2><p><strong>Overview</strong></p><p>MongoDB Atlas Vector Search is an advanced tool designed to handle complex vector similarity searches. It leverages the strengths of MongoDB's flexible data model and robust indexing capabilities, making it a powerful solution for various search and generative AI applications requiring vector search.</p><p><strong>Key benefits</strong></p><ol start="1"><li><p>Seamless integration with MongoDB: Atlas Vector Search is built into MongoDB, allowing you to use the same database for both structured and unstructured data. This integration simplifies your architecture and data management processes.</p></li><li><p>Scalability: MongoDB Atlas provides a highly scalable environment that can handle large volumes of data, making it ideal for applications requiring extensive vector searches.</p></li><li><p>Flexible indexing: MongoDB's indexing capabilities enable efficient storage and retrieval of vector data, ensuring fast and accurate search results.</p></li><li><p>Multi-cloud availability: Atlas Vector Search is available across major cloud providers, ensuring flexibility and reliability.</p></li><li><p>Security: Benefit from MongoDB's advanced security features, including encryption at rest and in transit, role-based access control, and comprehensive auditing.</p></li></ol><br /><h3 id="similarity-algorithms-supported">Similarity algorithms supported</h3><p>Cosine similarity: This measures the cosine of the angle between two vectors. It is particularly useful for comparing documents in text analysis, as it considers the orientation rather than the magnitude of the vectors.</p><p>Euclidean distance: This calculates the straight-line distance between two points in a multidimensional space. It is a simple and intuitive measure of similarity, often used in clustering and classification tasks.</p><p>Dot product: This computes the sum of the products of the corresponding entries of two sequences of numbers. It is used in various applications, including machine learning and recommendation systems, to measure the similarity between two vectors.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><div><h2 id="conclusion" style="font-family: Euclid Circular A; font-weight: 500;">Conclusion</h2><p>Vector embeddings represent a significant leap in how machines process and understand human language and other complex data types. From enhancing the capabilities of NLP in understanding text to their applications in fields like computer vision and bioinformatics, vector embeddings have proven to be invaluable tools. As technology evolves, so will the sophistication and utility of vector embeddings.</p><p>By storing vector embeddings in documents alongside metadata and contextual app data in a single, unified, fully managed, secure platform, developers can enjoy a seamless, flexible, and simplified experience. MongoDB’s robust integrations with all major AI services and cloud providers allow developers to use the embedding model of their choice and then perform indexing and searching, building apps efficiently and securely all in one place. This streamlined approach empowers developers to avoid the complexity of dealing with multiple platforms and focus more on building effective search and generative AI applications for their organizations. See how <a target="_self" href="/products/platform/atlas-vector-search">MongoDB Vector Search</a> works, and visit the Atlas Vector Search <a target="_target" href="https://www.mongodb.com/docs/atlas/atlas-vector-search/tutorials/vector-search-quick-start/?tck=ai_as_web">Quick Start guide</a> to create your first index in minutes.</p></div></div></div><div style="background: auto;" class="Article__ArticleStyles-sc-14w82yl-0 obHHn relative w-full "><div class="relative w-full w-max-770 p-h-15 p-v-25 m-auto"><h2 id="faqs" style="font-family: Euclid Circular A; font-weight: 500;">FAQs</h2></div></div><div class="relative"><div class="absolute w-full h-full overflow-hidden" style="z-index: -1;"><div style="background-position: center; background-repeat: no-repeat; background-size: cover; z-index: -1;" class="lazyload absolute w-full h-full"></div></div><div class="w-full w-max-770 m-auto p-h-20 p-t-10 p-b-10"><div class="Details__Container-sc-wfooue-0 emWlgh"><div class="fl-1"><h3 id="what-is-semantic-search" class="dark-green m-0 fnt-medium p-v-20">What is semantic search?</h3><div class="overflow-hidden" style="height: auto; max-height: 0; transition: 200ms ease 0ms; opacity: 0;"><div class="m-0 p-b-30 dark-gray fl fl-wrap"><p><a target="_target" href="https://www.mongodb.com/resources/basics/semantic-search">Semantic search</a> refers to a search technique that goes beyond keyword matching to understand the intent and contextual meaning of the search query. Instead of just looking for exact word matches, semantic search considers factors like the context of words in the query, the relationship between words, synonyms, and the overall meaning behind the query. This approach allows for more accurate and relevant search results, as it aligns more closely with how humans understand and use language.</p></div></div></div><button class="m-l-20 green fnt-24 p-h-20"><div style="transform: rotate(45deg); transition: 200ms ease 0ms;"><mdb-icon name="close_x" style="--icon-size: 15px;"></mdb-icon></div></button></div><div class="Details__Container-sc-wfooue-0 emWlgh"><div class="fl-1"><h3 id="does-a-reverse-image-search-involve-vector-embeddings" class="dark-green m-0 fnt-medium p-v-20">Does a reverse image search involve vector embeddings?</h3><div class="overflow-hidden" style="height: auto; max-height: 0; transition: 200ms ease 0ms; opacity: 0;"><div class="m-0 p-b-30 dark-gray fl fl-wrap"><p>Yes, in a reverse image search, images are transformed into vector embeddings, which are used to compare and find similar images in a database, making the search process efficient and accurate.</p></div></div></div><button class="m-l-20 green fnt-24 p-h-20"><div style="transform: rotate(45deg); transition: 200ms ease 0ms;"><mdb-icon name="close_x" style="--icon-size: 15px;"></mdb-icon></div></button></div><div class="Details__Container-sc-wfooue-0 emWlgh"><div class="fl-1"><h3 id="what-is-anomaly-detection" class="dark-green m-0 fnt-medium p-v-20">What is anomaly detection?</h3><div class="overflow-hidden" style="height: auto; max-height: 0; transition: 200ms ease 0ms; opacity: 0;"><div class="m-0 p-b-30 dark-gray fl fl-wrap"><p>Anomaly detection is a technique used in data analysis and various applications to identify patterns that do not conform to expected behavior. These nonconforming patterns are often referred to as anomalies, outliers, or exceptions.</p></div></div></div><button class="m-l-20 green fnt-24 p-h-20"><div style="transform: rotate(45deg); transition: 200ms ease 0ms;"><mdb-icon name="close_x" style="--icon-size: 15px;"></mdb-icon></div></button></div></div></div><style data-emotion="css-global wo4i12">@font-face{font-family:Akzidenz-Grotesk Std;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Regular-WebXL.woff2) format('woff2');font-weight:300;font-style:normal;font-display:swap;}@font-face{font-family:Akzidenz-Grotesk Std;src:url(https://static.mongodb.com/com/fonts/EuclidCircularA-Medium-WebXL.woff2) format('woff2');font-weight:500;font-style:normal;font-display:swap;}</style><style data-emotion="css 1j19lrv">.css-1j19lrv{width:100%;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;background-color:#061621;color:#ffffff;}.css-1j19lrv a:hover{-webkit-text-decoration:underline;text-decoration:underline;}</style><footer class="css-1j19lrv"><style data-emotion="css 1el4ear">.css-1el4ear{width:100%;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;-webkit-box-flex-wrap:wrap;-webkit-flex-wrap:wrap;-ms-flex-wrap:wrap;flex-wrap:wrap;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;max-width:1024px;padding-top:64px;padding-bottom:64px;padding-left:24px;padding-right:24px;}@media screen and (min-width: 768px){.css-1el4ear{-webkit-box-flex-wrap:unset;-webkit-flex-wrap:unset;-ms-flex-wrap:unset;flex-wrap:unset;}}</style><style data-emotion="css 1ro0ep5">.css-1ro0ep5{box-sizing:border-box;margin:0;min-width:0;width:100%;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;-webkit-box-flex-wrap:wrap;-webkit-flex-wrap:wrap;-ms-flex-wrap:wrap;flex-wrap:wrap;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;max-width:1024px;padding-top:64px;padding-bottom:64px;padding-left:24px;padding-right:24px;}@media screen and (min-width: 768px){.css-1ro0ep5{-webkit-box-flex-wrap:unset;-webkit-flex-wrap:unset;-ms-flex-wrap:unset;flex-wrap:unset;}}</style><div class="css-1ro0ep5"><style data-emotion="css 1dkb3g4">.css-1dkb3g4{-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;min-width:100%;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;margin-bottom:40px;}@media screen and (min-width: 768px){.css-1dkb3g4{min-width:unset;margin-bottom:0px;}}</style><style data-emotion="css dskpio">.css-dskpio{box-sizing:border-box;margin:0;min-width:0;-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;min-width:100%;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;margin-bottom:40px;}@media screen and (min-width: 768px){.css-dskpio{min-width:unset;margin-bottom:0px;}}</style><div class="css-dskpio"><style data-emotion="css sbarov">.css-sbarov{display:grid;grid-template-columns:1fr 1fr;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;}@media screen and (min-width: 768px){.css-sbarov{display:block;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-align-items:flex-start;-webkit-box-align:flex-start;-ms-flex-align:flex-start;align-items:flex-start;}}</style><style data-emotion="css tucx3n">.css-tucx3n{box-sizing:border-box;margin:0;min-width:0;display:grid;grid-template-columns:1fr 1fr;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;-webkit-box-pack:justify;-webkit-justify-content:space-between;justify-content:space-between;}@media screen and (min-width: 768px){.css-tucx3n{display:block;-webkit-flex-direction:column;-ms-flex-direction:column;flex-direction:column;-webkit-align-items:flex-start;-webkit-box-align:flex-start;-ms-flex-align:flex-start;align-items:flex-start;}}</style><div class="css-tucx3n"><style data-emotion="css 1daqsee">.css-1daqsee{max-width:120px;}@media screen and (min-width: 1024px){.css-1daqsee{max-width:150px;}}</style><style data-emotion="css 1a7nq3w">.css-1a7nq3w{box-sizing:border-box;margin:0;min-width:0;max-width:120px;}@media screen and (min-width: 1024px){.css-1a7nq3w{max-width:150px;}}</style><div class="css-1a7nq3w"><style data-emotion="css 93rpjy">.css-93rpjy{font-size:14px;line-height:18px;height:auto;}</style><a href="https://www.mongodb.com" class="css-93rpjy"><style data-emotion="css 1xgosv1">.css-1xgosv1{width:100%;min-width:100px;font-size:14px;line-height:18px;height:auto;}</style><img src="https://webimages.mongodb.com/_com_assets/cms/kuyj3d95v5vbmm2f4-horizontal_white.svg?auto=format%252Ccompress" alt="MongoDB logo" width="150" height="38" class="css-1xgosv1" /></a></div></div><style data-emotion="css yjv4kt">.css-yjv4kt{display:none;font-family:Akzidenz-Grotesk Std;font-weight:300;font-size:12px;color:#b8c4c2;}@media screen and (min-width: 768px){.css-yjv4kt{display:block;}}</style><style data-emotion="css 148f6w8">.css-148f6w8{box-sizing:border-box;margin:0;min-width:0;display:none;font-family:Akzidenz-Grotesk Std;font-weight:300;font-size:12px;color:#b8c4c2;}@media screen and (min-width: 768px){.css-148f6w8{display:block;}}</style><div class="css-148f6w8">© 2024 MongoDB, Inc.</div></div><style data-emotion="css 1svoxwz">.css-1svoxwz{-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;margin-top:24px;min-width:50%;}@media screen and (min-width: 460px){.css-1svoxwz{min-width:unset;}}@media screen and (min-width: 768px){.css-1svoxwz{margin-top:0px;min-width:unset;}}</style><style data-emotion="css 1err1dc">.css-1err1dc{box-sizing:border-box;margin:0;min-width:0;-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;margin-top:24px;min-width:50%;}@media screen and (min-width: 460px){.css-1err1dc{min-width:unset;}}@media screen and (min-width: 768px){.css-1err1dc{margin-top:0px;min-width:unset;}}</style><div class="css-1err1dc"><style data-emotion="css of25nb">.css-of25nb{font-family:Akzidenz-Grotesk Std;font-weight:500;font-size:16px;margin-bottom:24px;color:#ffffff;margin-top:0px;}@media screen and (min-width: 768px){.css-of25nb{margin-top:initial;}}</style><p class="css-of25nb">About</p><style data-emotion="css 1akr5ww">.css-1akr5ww{list-style:none;margin:0;padding:0;display:block;}</style><ul class="css-1akr5ww"><style data-emotion="css 1w8osvb">.css-1w8osvb{margin-bottom:24px;}</style><li class="css-1w8osvb"><style data-emotion="css 16ay36s">.css-16ay36s{color:#ffffff;-webkit-text-decoration:none;text-decoration:none;font-family:Akzidenz-Grotesk Std;font-weight:300;font-size:14px;line-height:32px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}</style><a href="https://www.mongodb.com/careers" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Careers</a></li><li class="css-1w8osvb"><a href="https://investors.mongodb.com" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Investor Relations</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/legal" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Legal</a></li><li class="css-1w8osvb"><a href="https://github.com/mongodb" target="_self" rel="noreferrer noopener" class=" css-16ay36s">GitHub</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/company/contact/mongodb-vulnerability-disclosure-policy" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Security Information</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/products/platform/trust" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Trust Center</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/company/contact/social-media-hub" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Connect with Us</a></li></ul></div><div class="css-1err1dc"><p class="css-of25nb">Support</p><ul class="css-1akr5ww"><li class="css-1w8osvb"><a href="https://www.mongodb.com/company/contact" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Contact Us</a></li><li class="css-1w8osvb"><a href="https://support.mongodb.com/welcome" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Customer Portal</a></li><li class="css-1w8osvb"><a href="https://status.mongodb.com/" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Atlas Status</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/services/support" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Customer Support</a></li></ul></div><div class="css-1err1dc"><p class="css-of25nb">Deployment Options</p><ul class="css-1akr5ww"><li class="css-1w8osvb"><a href="https://www.mongodb.com/cloud/atlas/register" target="_self" rel="noreferrer noopener" class=" css-16ay36s">MongoDB Atlas</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/try/download/enterprise" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Enterprise Advanced</a></li><li class="css-1w8osvb"><a href=" https://www.mongodb.com/try/download/community" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Community Edition</a></li></ul></div><style data-emotion="css 17c2rye">.css-17c2rye{-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;margin-top:24px;min-width:50%;}@media screen and (min-width: 460px){.css-17c2rye{min-width:100%;}}@media screen and (min-width: 768px){.css-17c2rye{margin-top:0px;min-width:unset;}}</style><style data-emotion="css 130l0sz">.css-130l0sz{box-sizing:border-box;margin:0;min-width:0;-webkit-flex-basis:25%;-ms-flex-preferred-size:25%;flex-basis:25%;margin-top:24px;min-width:50%;}@media screen and (min-width: 460px){.css-130l0sz{min-width:100%;}}@media screen and (min-width: 768px){.css-130l0sz{margin-top:0px;min-width:unset;}}</style><div class="css-130l0sz"><p class="css-of25nb">Data Basics</p><style data-emotion="css ro7s69">.css-ro7s69{list-style:none;margin:0;padding:0;display:block;grid-template-columns:1fr 1fr 1fr;}@media screen and (min-width: 460px){.css-ro7s69{display:grid;}}@media screen and (min-width: 768px){.css-ro7s69{display:block;}}</style><ul class="css-ro7s69"><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/basics/databases/vector-databases" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Vector Databases</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/basics/databases/nosql-explained" target="_self" rel="noreferrer noopener" class=" css-16ay36s">NoSQL Databases</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/basics/databases/document-databases" target="_self" rel="noreferrer noopener" class=" css-16ay36s">Document Databases</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/basics/artificial-intelligence/retrieval-augmented-generation" target="_self" rel="noreferrer noopener" class=" css-16ay36s">RAG Database</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/basics/databases/acid-transactions" target="_self" rel="noreferrer noopener" class=" css-16ay36s">ACID Transactions</a></li><li class="css-1w8osvb"><a href="https://www.mongodb.com/resources/languages/mern-stack" target="_self" rel="noreferrer noopener" class=" css-16ay36s">MERN Stack</a></li><li class="css-1w8osvb"><a href=" https://www.mongodb.com/resources/languages/mean-stack" target="_self" rel="noreferrer noopener" class=" css-16ay36s">MEAN Stack</a></li></ul></div><style data-emotion="css 9xnpax">.css-9xnpax{display:block;font-family:Akzidenz-Grotesk Std;width:100%;font-weight:300;font-size:12px;color:#b8c4c2;margin-top:24px;text-align:center;}@media screen and (min-width: 768px){.css-9xnpax{display:none;}}</style><style data-emotion="css 14jzdz8">.css-14jzdz8{box-sizing:border-box;margin:0;min-width:0;display:block;font-family:Akzidenz-Grotesk Std;width:100%;font-weight:300;font-size:12px;color:#b8c4c2;margin-top:24px;text-align:center;}@media screen and (min-width: 768px){.css-14jzdz8{display:none;}}</style><div class="css-14jzdz8">© 2024 MongoDB, Inc.</div></div></footer></div></div></div> <script id="server-data"> window.__serverData={"_id":"66885d2e687f6d6235b699fd","url":"resources/basics/vector-embeddings","components":[{"key":"Nav","props":{"left":[{"title":"Cloud","links":[{"title":"Atlas","text":"Fully managed cloud database","href":"/cloud/atlas"},{"title":"Atlas Data Lake","text":"Query and combine AWS S3 and MongoDB data","href":"/atlas/data-lake"},{"title":"Atlas Search","text":"Cloud-native full-text search engine","href":"/atlas/search"},{"title":"Realm","text":"Application Development Services","href":"/realm"},{"title":"Charts","text":"Native visualization for MongoDB data","href":"/products/charts"},{"title":"Atlas for Government","text":"Atlas for US Government workloads","href":"/cloud/atlas/government"}]},{"title":"Software","links":[{"title":"Community Server","text":"A free and open document database","href":"/try/download/community"},{"title":"Enterprise Server","text":"Advanced features and security","href":"/try/download/enterprise"},{"title":"Developer Tools","text":"Connect, configure and work with MongoDB","href":"/developer-tools"},{"title":"Compass","text":"GUI for MongoDB","href":"/products/compass"},{"title":"Ops Manager","text":"On-prem management platform for MongoDB","href":"/products/ops-manager"},{"title":"Connectors","text":"Easy integrations to your data estate","href":"/connectors"}]},{"title":"Pricing","links":[],"href":"/pricing"},{"title":"Learn","links":[{"title":"What is MongoDB?","text":"Start here","href":"/what-is-mongodb"},{"title":"University","href":"https://university.mongodb.com","text":"Free online courses from MongoDB"},{"title":"Blog","href":"/blog","text":"Updates, tutorials, people"},{"title":"Developer Hub","href":"https://developer.mongodb.com","text":"Developer best practices, trends, insights"},{"title":"Resources","href":"/resources","text":"Webinars, white papers, datasheets, and more"},{"title":"Training","href":"/training","text":"Instructor-led sessions on your schedule"},{"title":"Events","href":"/events","text":"Worldwide community events"},{"title":"Community","href":"https://community.mongodb.com","text":"The MongoDB Community discussion forums"}]},{"title":"Solutions","links":[{"title":"Customers","text":"Who uses MongoDB","href":"/who-uses-mongodb"},{"title":"Use Cases","text":"How MongoDB is used","href":"/use-cases"},{"title":"Consulting","text":"Accelerate success with MongoDB","href":"/products/consulting"},{"title":"Partners","text":"Find or become a partner","href":"/partners"}]},{"title":"Docs","links":[{"title":"Cloud","text":"Atlas, Realm, and more","href":"https://docs.mongodb.com/cloud/"},{"title":"Server","href":"https://docs.mongodb.com/manual/","text":"The database"},{"title":"Drivers","text":"Language APIs","href":"https://docs.mongodb.com/ecosystem/drivers/"},{"title":"Tools","text":"Compass, Charts, Connectors, and more","href":"https://docs.mongodb.com/tools/"},{"title":"How to Guides","text":"Get started in minutes","href":"https://docs.mongodb.com/guides/"}]}],"right":[{"title":"Contact","href":"/contact","button":false},{"title":"Sign In","href":"https://cloud.mongodb.com/user","button":false},{"title":"Try Free","href":"/try","button":true}],"mobile":[{"text":"Contact","href":"/contact"}],"navType":"","notSticky":false,"banner":{"bannerTheme":"","pillText":"","bannerText":"","href":""},"isATF":true},"id":"f93b19ee-9a15-44b6-9563-f309b7d880f0"},{"key":"Header","props":{"h1":"What are Vector Embeddings? ","titleColor":0,"content":"","cta":{"href":"https://www.mongodb.com/cloud/atlas/register","text":"Get started free","openInNewWindow":true,"faux":false},"cta2":{"href":"","text":"","openInNewWindow":false,"faux":false},"imgsrc":"https://webassets.mongodb.com/_com_assets/cms/header-au039oiqsr.svg","bgColor":"","bgImage":"","isATF":true},"id":1720212972605},{"key":"Article","props":{"content":"Vector embeddings are mathematical representations of text created by translating words or sentences into numbers — a language that computers can understand. They bridge the rich, nuanced world of human language (text, images, speeches, videos etc.) and the precise environment of machine learning models (numbers) by representing data points.\n\n![An image of vector embeddings including unstructured data and encoder.](https://webimages.mongodb.com/_com_assets/cms/m3n5dd9oq8zrgpot7-vector-embedding.png?auto=format%252Ccompress)\n\n<br>\n\nMost often used in natural language processing (NLP), vector embeddings allow machine learning algorithms to analyze information much like humans but at a scale and speed far beyond our capabilities. \nAlthough they also work with images, audio processing, bioinformatics, and recommendation systems, this article will focus on word vector embeddings in natural language processing using a machine learning model. \n\n<br>\n\n**Table of contents**\n\n- [What is natural language processing?](#what-is-natural-language-processing)\n- [Types of vector embeddings in NLP](#types-of-vector-embeddings-in-nlp)\n- [What is a vector?](#what-is-a-vector-)\n- [Vector embeddings in multidimensional spaces](#vector-embeddings-in-multidimensional-spaces)\n- [Dimensionality in vector embeddings](#dimensionality-in-vector-embeddings)\n- [Advantages of vector embeddings in real-world scenarios](#advantages-of-vector-embeddings-in-realworld-scenarios)\n- [Challenges and limitations](#challenges-and-limitations)\n- [Using vector embeddings in other applications](#using-vector-embeddings-in-other-applications-)\n- [Why use MongoDB Atlas Vector Search for vector similarity search?](#why-use-mongodb-atlas-vector-search-for-vector-similarity-search)\n- [Conclusion](#conclusion)\n- [FAQs](#faqs)\n\n","textalign":"","background":"","isATF":true},"id":1720212947522},{"key":"Article","props":{"content":"##What is natural language processing?\nNLP is a type of artificial intelligence that uses vector embeddings in conjunction with machine learning algorithms to evaluate, understand, and interpret human language. This combination achieves comprehension and interaction that mirrors human ability but at a scale and speed far beyond our capabilities. NLP excels in tasks such as interpreting text from social media, translating languages, and powering conversational agents.\n","textalign":"","background":"","isATF":false},"id":1720213124707},{"key":"Article","props":{"content":"##Types of vector embeddings in NLP\nBelow are a few examples of the diverse vector embedding techniques instrumental in advancing NLP, each bringing its strengths to various language understanding challenges.\n\n- **Word2Vec**: Developed by Google, Word2Vec captures the context of words within documents. It’s beneficial for tasks that require understanding word associations and meanings based on their usage in sentences.\n\n- **GloVe (global vectors for word representation)**: GloVe is unique in its approach as it analyzes word co-occurrences over the whole corpus for training, enabling it to capture global statistics of words. It’s particularly useful for tasks that involve semantic similarity between words.\n\n- **BERT (Bidirectional Encoder Representations from Transformers)**: Developed by Google, BERT represents a breakthrough in contextually aware embeddings. It looks at the context from both sides of a word in a sentence, making it highly effective for sophisticated tasks like sentiment analysis and question answering.","textalign":"","background":"","isATF":false},"id":1720213145679},{"key":"Article","props":{"content":"##What is a vector? \nTo truly appreciate what vector embeddings are and how they work, it's essential first to understand what a vector is in this setting. Think of a vector as a point in space with direction and magnitude. It's like a dot on a map with specific coordinates. These data points aren't just random numbers; they represent different characteristics or features of the data types that the vector represents. \n","textalign":"","background":"","isATF":false},"id":1720213160578},{"key":"Article","props":{"content":"##Vector embeddings in multidimensional spaces\nNow that the basics of vectors and data points have been introduced, it's important to note that in the context of vector representations for text, more than one vector embedding or one dimension is used. When text — words, phrases, or entire documents — are converted into vectors, each piece of text plots as a point in a vast, multidimensional space. This space isn't like a typical three-dimensional space we're familiar with; it has many more dimensions, each representing a different aspect of the text's meaning or usage. \n\nImagine a map where words with similar meanings or usages are mapped closer together, making it easier to see their relationships or similar data points.\n\nTurning text into vector embeddings is a game-changer for machine learning algorithms. These algorithms are great at dealing with numbers — they can spot patterns, make comparisons, and draw conclusions from numerical data.\n","textalign":"","background":"","isATF":false},"id":1720213183671},{"key":"Article","props":{"content":"##Dimensionality in vector embeddings\nLet's dig a little deeper into the concept of dimensionality, which we introduced above. Think of dimensionality in vector embeddings like the resolution of a photo. High-resolution photos are more detailed and precise, but they take up more space on your phone and require more processing power. Similarly, in vector embeddings, more dimensions mean that the representation of words or phrases can capture more details and nuances of language.\n\n<br>\n\n###High-dimensional embeddings\nHigh-dimensional embeddings are like high-resolution photos. They have hundreds or even thousands of dimensions, allowing them to capture much information about a word or phrase. Each dimension can represent a different aspect of a word's meaning or use. This detailed representation is excellent for complex tasks in natural language processing, where understanding subtle differences in language is crucial. \n\nHowever, like high-resolution photos, these embeddings require more computer memory and processing power. Also, there's a risk of \"overfitting\" — think of it like a camera that focuses on capturing every tiny detail and fails to recognize common, everyday objects. In machine learning, the model might get too tailored to its training data and perform poorly on new, unseen data.\n\n<br>\n\n###Low-dimensional embeddings\nOn the other hand, low-dimensional embeddings are like lower-resolution photos. They have fewer dimensions, so they use less computer memory and process more quickly, which is excellent for applications that need to run fast or have limited resources. But just like lower-resolution photos can miss finer details, these embeddings might not capture all the subtle nuances of language. Depending on the task, they provide a more general picture, which can sometimes be enough.\n\nChoosing the proper dimensionality for creating vector embeddings is a balance. It's about weighing the need for detail against the need for efficiency and the ability of the model to perform well on new, unseen data. Finding the right balance often involves trial and error and depends on the specific task and the data. It's a crucial part of developing effective NLP solutions, requiring a thoughtful approach to meet both the linguistic needs of the task and the practical limitations of technology.\n","textalign":"","background":"","isATF":false},"id":1720213190910},{"key":"Article","props":{"content":"##Advantages of vector embeddings in real-world scenarios\nVector embeddings have opened up a world of possibilities in how machines interact with human language. They make technology more intuitive and natural, enriching interactions across digital platforms and tools. Below are a few applications highlighting how vector embeddings are used today. \n\n<br>\n\n###Sentiment analysis\nSentiment analysis is like a digital mood ring. Businesses use it to understand how people feel about their products or services by analyzing the tone of customer reviews and social media posts. Vector embeddings help computers catch subtle emotional cues in text, distinguishing genuine praise from sarcasm, even when the words are similar.\n\n<br>\n\n###Machine translation\nVector embeddings are the backbone of translation apps. They help computers grasp the complexities and nuances of different languages. When a sentence is translated from one language to another, it's not just about swapping words; it's about conveying the same meaning, tone, and context. Vector embeddings are crucial in achieving this.\n\n<br>\n\n###Chatbots and virtual assistants\nAre you curious how virtual assistants like Siri or Alexa understand and respond to your queries so well? This functionality is primarily due to vector embeddings. They enable artificial intelligence (AI) systems to process what you're saying, figure out what you mean, and respond in a way that makes sense.\n\n<br>\n\n###Information retrieval\nThis category covers everything from search engines to recommendation systems. Vector embeddings help these systems understand what is being searched, not just by matching keywords but by grasping the context of the query. This way, the information or recommendations will more likely be relevant.\n\n<br>\n\n###Text classification\nText classification can filter emails, categorize news articles, and even tag social media posts. Vector embeddings assist in sorting text into different categories by understanding the underlying themes and topics, making it easier for algorithms to decide, for example, if an email is spam.\n\n<br>\n\n###Speech recognition\nRegarding converting spoken words into written text, vector embeddings play a crucial role. They help capture the spoken words accurately, considering how the same word can be pronounced or used in different contexts, leading to more accurate transcriptions.\n","textalign":"","background":"","isATF":false},"id":1720213253255},{"key":"Article","props":{"content":"##Challenges and limitations\nWhile vector embeddings are a powerful tool in NLP, they are not without their challenges. Addressing these issues is crucial for ensuring that these technologies are effective, fair, and up-to-date, requiring continuous effort and innovation in the field. Let's explore these challenges and limitations, especially in how vector embeddings interact with and process human language.\n\n<br>\n\n###Handling out-of-vocabulary words\nOne of the trickiest issues with vector embeddings is dealing with words that the system has never seen before, often called \"out-of-vocabulary\" words. It's like encountering a word in a foreign language you've never learned. For a computer, these new words can be a fundamental stumbling block. The system might struggle to understand and place them correctly in the context of what it already knows.\n\n<br>\n\n###Bias in embeddings\nLike humans, computers can be biased, especially when they learn from data reflecting human prejudices. When vector embeddings train on text data from the internet or other human-generated sources, they can inadvertently pick up and even amplify these biases. This possibility is significant because it can lead to unfair or stereotypical representations in various applications, like search engines or AI assistants.\n\n<br>\n\n###Complexity of maintaining and updating models\nKeeping vector embedding models up-to-date and relevant is no small task. Language is constantly evolving — new words pop up, old words fade away, and meanings change. Ensuring these models stay current is like updating a constantly evolving dictionary. It requires ongoing work and resources, making it a complex and challenging aspect of working with vector embeddings.\n\n<br>\n\n###Contextual ambiguity\nWhile vector embeddings are good at capturing meaning, they sometimes struggle with words with multiple meanings based on context. For instance, the word \"bat\" can refer to an animal or sports equipment, and without sufficient context, the model might not accurately capture the intended use.\n\n<br>\n\n###Resource intensity\nTraining sophisticated vector embedding models requires significant computational resources, which can be a barrier, especially for smaller organizations or individual researchers who might not have access to the necessary computing power.\n\n<br>\n\n###Data quality and availability\nThe effectiveness of vector embeddings heavily depends on the quality and quantity of the training data. The embeddings might not be as accurate or helpful in languages or domains where data is scarce or of poor quality.\n\n<br>\n\n###Transferability across languages\nVector embeddings trained in one language may not transfer well to another, especially for structurally different languages. These structural differences challenge multilingual applications or languages with limited resources.\n\n<br>\n###Model interpretability\nUnderstanding why a vector embedding model behaves a certain way or makes specific decisions can be challenging. This lack of interpretability can be a significant issue, especially in applications where understanding the model's reasoning is crucial.\n\n<br>\n\n###Scalability\nAs the amount of data and the complexity of tasks increase, scaling vector embedding models while maintaining performance and efficiency can be challenging.\n\n<br>\n\n###Dependency on training data\n\nVector embeddings can only be as good as their training data. If the training data is limited or biased, the embeddings will inherently reflect those limitations or biases.\n","textalign":"","background":"","isATF":false},"id":1720213273484},{"key":"Article","props":{"content":"##Using vector embeddings in other applications \nWhile vector embeddings are most prominently used in NLP, they are also used in other ways. Here are a few other areas where vector embeddings are employed. \n\n<br>\n\n###Computer vision\nIn image processing and computer vision, embeddings represent images or parts of images. Similar to how they capture the essence of words in NLP, embeddings in computer vision capture essential features of images, enabling tasks like image recognition, classification, and similarity detection.\n\n<br>\n\n###Recommendation systems\nVector embeddings also show up in recommendation systems, such as those on e-commerce or streaming platforms. They help understand user preferences and item characteristics by representing users and items in a vector space, enabling the system to make personalized recommendations based on similarity.\n\n<br>\n\n###Bioinformatics\nIn bioinformatics, embeddings can represent biological data, such as gene sequences or protein structures. These embeddings help in various predictive tasks, like understanding gene function or protein-protein interactions.\n\n<br>\n\n###Graph analysis\nIn network and graph analysis, embeddings represent nodes and edges of a graph, which is helpful in social network analysis, link prediction, and understanding the structure and dynamics of complex systems.\n\n<br>\n\n###Time series analysis\nVector embeddings work in analyzing time series data, such as financial market trends or sensor data, by capturing temporal patterns and dependencies in a vector space.\n\nThese diverse applications show that the concept of embeddings is a versatile tool in the broader field of machine learning and data science, not limited to just text and language processing.\n","textalign":"","background":"","isATF":false},"id":1720213303077},{"key":"Article","props":{"content":"##Why use MongoDB Atlas Vector Search for vector similarity search?\n\n**Overview**\n\nMongoDB Atlas Vector Search is an advanced tool designed to handle complex vector similarity searches. It leverages the strengths of MongoDB's flexible data model and robust indexing capabilities, making it a powerful solution for various search and generative AI applications requiring vector search.\n\n**Key benefits**\n\n1. Seamless integration with MongoDB: Atlas Vector Search is built into MongoDB, allowing you to use the same database for both structured and unstructured data. This integration simplifies your architecture and data management processes.\n\n2. Scalability: MongoDB Atlas provides a highly scalable environment that can handle large volumes of data, making it ideal for applications requiring extensive vector searches.\n\n2. Flexible indexing: MongoDB's indexing capabilities enable efficient storage and retrieval of vector data, ensuring fast and accurate search results.\n\n4. Multi-cloud availability: Atlas Vector Search is available across major cloud providers, ensuring flexibility and reliability.\n\n5. Security: Benefit from MongoDB's advanced security features, including encryption at rest and in transit, role-based access control, and comprehensive auditing.\n\n<br>\n\n###Similarity algorithms supported\nCosine similarity: This measures the cosine of the angle between two vectors. It is particularly useful for comparing documents in text analysis, as it considers the orientation rather than the magnitude of the vectors.\n\nEuclidean distance: This calculates the straight-line distance between two points in a multidimensional space. It is a simple and intuitive measure of similarity, often used in clustering and classification tasks.\n\nDot product: This computes the sum of the products of the corresponding entries of two sequences of numbers. It is used in various applications, including machine learning and recommendation systems, to measure the similarity between two vectors.\n","textalign":"","background":"","isATF":false},"id":1720213320309},{"key":"Article","props":{"content":"##Conclusion\nVector embeddings represent a significant leap in how machines process and understand human language and other complex data types. From enhancing the capabilities of NLP in understanding text to their applications in fields like computer vision and bioinformatics, vector embeddings have proven to be invaluable tools. As technology evolves, so will the sophistication and utility of vector embeddings.\n\nBy storing vector embeddings in documents alongside metadata and contextual app data in a single, unified, fully managed, secure platform, developers can enjoy a seamless, flexible, and simplified experience. MongoDB’s robust integrations with all major AI services and cloud providers allow developers to use the embedding model of their choice and then perform indexing and searching, building apps efficiently and securely all in one place. This streamlined approach empowers developers to avoid the complexity of dealing with multiple platforms and focus more on building effective search and generative AI applications for their organizations. See how [MongoDB Vector Search](/products/platform/atlas-vector-search) works, and visit the Atlas Vector Search [Quick Start guide](https://www.mongodb.com/docs/atlas/atlas-vector-search/tutorials/vector-search-quick-start/?tck=ai_as_web) to create your first index in minutes.","textalign":"","background":"","isATF":false},"id":1720213343098},{"key":"Article","props":{"content":"##FAQs","textalign":"","background":"","isATF":false},"id":1720213371255},{"key":"Details","props":{"items":[{"titleColor":0,"title":"What is semantic search?","text":"\n[Semantic search](https://www.mongodb.com/resources/basics/semantic-search) refers to a search technique that goes beyond keyword matching to understand the intent and contextual meaning of the search query. Instead of just looking for exact word matches, semantic search considers factors like the context of words in the query, the relationship between words, synonyms, and the overall meaning behind the query. This approach allows for more accurate and relevant search results, as it aligns more closely with how humans understand and use language."},{"titleColor":0,"title":"Does a reverse image search involve vector embeddings?","text":"\nYes, in a reverse image search, images are transformed into vector embeddings, which are used to compare and find similar images in a database, making the search process efficient and accurate."},{"titleColor":0,"title":"What is anomaly detection?","text":"\nAnomaly detection is a technique used in data analysis and various applications to identify patterns that do not conform to expected behavior. These nonconforming patterns are often referred to as anomalies, outliers, or exceptions."}],"bgColor":"","bgImage":"","isATF":false},"id":1720213389786},{"key":"Footer","props":{"toggle":0,"column1":{"title":"Resources","maxWidth":"185","hasIcons":0,"className":"","items":[{"href":"/nosql-explained","text":"NoSQL Database Explained","isTarget":""},{"href":"/collateral/mongodb-architecture-guide","text":"MongoDB Architecture Guide","isTarget":""},{"href":"/products/mongodb-enterprise-advanced","text":"MongoDB Enterprise Advanced","isTarget":""},{"href":"/cloud/atlas","text":"MongoDB Atlas","isTarget":""},{"href":"/cloud/stitch","text":"MongoDB Stitch","isTarget":""},{"href":"//engineering.mongodb.com/","text":"MongoDB Engineering Blog","isTarget":"true"}]},"column2":{"title":"Education & Support","maxWidth":"150","hasIcons":0,"className":"","items":[{"href":"//university.mongodb.com/courses/catalog","text":"View Course Catalog","isTarget":"true"},{"href":"//university.mongodb.com/certification","text":"Certification","isTarget":"true"},{"href":"//docs.mongodb.com/manual/","text":"MongoDB Manual","isTarget":"true"},{"href":"//docs.mongodb.com/manual/installation/","text":"Installation","isTarget":"true"},{"href":"//support.mongodb.com/welcome","text":"Support","isTarget":""},{"href":"/faq","text":"FAQ","isTarget":""}]},"column3":{"title":"Popular Topics","maxWidth":"300","hasIcons":0,"className":"be-ix-link-block","items":[{"href":"/cloud/atlas/aws-mongodb","text":"MongoDB on AWS","isTarget":""},{"href":"/cloud/atlas/mongodb-google-cloud","text":"MongoDB on Google Cloud","isTarget":""},{"href":"/cloud/atlas/multicloud-data-distribution","text":"Run MongoDB on Multiple Clouds with MongoDB Atlas","isTarget":""},{"href":"/cloud/atlas/migrate","text":"Migrate to MongoDB Atlas","isTarget":""},{"href":"/cloud-database","text":"What is a Cloud Database?","isTarget":""},{"href":"/blog/post/building-a-rest-api-with-mongodb-stitch","text":"Building a REST API with MongoDB Stitch","isTarget":""}]},"column4":{"title":"About","maxWidth":"100","hasIcons":0,"className":"","items":[{"href":"/company","text":"MongoDB, Inc.","isTarget":""},{"href":"/leadership","text":"Leadership","isTarget":""},{"href":"/company/newsroom","text":"Press Room","isTarget":""},{"href":"/careers","text":"Careers","isTarget":""},{"href":"https://investors.mongodb.com","text":"Investors","isTarget":""},{"href":"/legal/legal-notices","text":"Legal Notices","isTarget":""},{"href":"/legal/privacy-policy","text":"Privacy Notice","isTarget":""},{"href":"/security","text":"Security Information","isTarget":""},{"href":"/cloud/trust","text":"Trust Center","isTarget":""},{"href":"/office-locations","text":"Office Locations","isTarget":""},{"href":"/community-code-of-conduct","text":"Code of Conduct","isTarget":""}]},"column5":{"title":"Follow Us","maxWidth":"120","hasIcons":1,"className":"","items":[{"href":"//facebook.com/mongodb","text":"Facebook","isTarget":"true"},{"href":"//github.com/mongodb","text":"Github","isTarget":"true"},{"href":"//youtube.com/user/mongodb","text":"Youtube","isTarget":"true"},{"href":"//twitter.com/mongodb","text":"Twitter","isTarget":"true"},{"href":"//www.linkedin.com/company/mongodbinc/","text":"LinkedIn","isTarget":"true"},{"href":"//slackpass.io/mongo-db","text":"Slack","isTarget":"true"},{"href":"//stackoverflow.com/tags/mongodb/info","text":"StackOverflow","isTarget":"true"}]},"isATF":false},"id":"aecfeaf5-65f6-4623-ad36-94949d7aef26"}],"created_at":"2024-07-05T20:53:02.596Z","meta":{"title":"What Are Vector Embeddings? | MongoDB","title#localised":true,"description":"Learn the basics of vector embeddings, its role in AI, and how MongoDB utilizes this technology.","description#localised":true},"node_type":"content_block","owners":[],"published_at":"2024-07-08T16:45:21.072Z","status":"published","updated_at":"2024-11-24T12:58:55.262Z","cms":{"editedURL":true},"draft":true,"globals":[{"_id":"6001f22ac1f95e773a0e0044","key":"AccountLogin","created_at":"2021-01-15T19:51:06.717Z","props":{"title":"MongoDB Stands with the Black Community, changes","subtitle":"Join MongoDB in supporting organizations that are fighting for racial justice and equal opportunity","cta":{"text":"Join Now","href":"https://mongodbforjustice.mongodb.events/","openInNewWindow":false,"faux":false},"image":{"desktop":"https://account.mongodb.com/static/images/auth/racial_justice_desktop_login.png","mobile":"https://account.mongodb.com/static/images/auth/racial_justice_mobile.png"},"artist":"Artwork by [Lo Harris](http://loharris.com/)"},"updated_at":"2024-10-21T20:11:05.302Z"},{"_id":"601c7536f53e6b3af09679d3","key":"PromoBanner","created_at":"2021-02-04T22:29:10.420Z","props":{"type":4,"typeColor":0,"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","titleColor":0,"background":1,"disabled":false,"eventBranded":false,"eventBrandedButtonImg":0},"updated_at":"2024-10-21T20:11:05.312Z","translations":{"en-us":{"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","eventBranded":false,"disabled":false,"type":4,"eventBrandedButtonImg":0,"typeColor":0,"background":1,"titleColor":0},"pt-br":{"title":"Register for MongoDB.live today!","type":0,"titleColor":0},"es":{"title":"Register for MongoDB.live today!"},"it-it":{"title":"Register for MongoDB.live today!"},"de-de":{"title":"Register for MongoDB.live today!"},"fr-fr":{"title":"Register for MongoDB.live today!"},"ja-jp":{"title":"Register for MongoDB.live today!"},"ko-kr":{"title":"Register for MongoDB.live today!"},"zh-cn":{"title":"Register for MongoDB.live today!"}}},{"_id":"60c127b5527761a42edca7bb","key":"TranslationFallbackBanner","created_at":"2021-06-09T20:42:29.953Z","updated_at":"2024-10-21T20:11:05.319Z","props":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"translations":{"en-us":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"pt-br":{"text":"O conteúdo desta página não está disponível no idioma selecionado no momento. No entanto, estamos comprometidos em oferecer o máximo de conteúdo localizado possível. Agradecemos a paciência."},"es":{"text":"El contenido de esta página no está disponible actualmente en el idioma seleccionado. Sin embargo, nos comprometemos a proporcionar la mayor cantidad de contenido localizado posible. Gracias por tu paciencia."},"it-it":{"text":"I contenuti di questa pagina non sono attualmente disponibili nella lingua selezionata. Tuttavia, ci impegniamo a fornire il maggior numero possibile di contenuti localizzati. Grazie per la pazienza."},"de-de":{"text":"Die Inhalte dieser Seite sind derzeit nicht in der gewählten Sprache verfügbar. Wir arbeiten jedoch daran, so viele lokalisierte Inhalte wie möglich bereitzustellen. Vielen Dank für Ihre Geduld."},"fr-fr":{"text":"Le contenu de cette page n'est actuellement pas disponible dans la langue sélectionnée. Nous mettons toutefois tout en œuvre pour proposer autant de contenu localisé que possible. Merci de votre patience."},"ja-jp":{"text":"現在、このページの選択した言語のコンテンツはありません。ローカライズされたコンテンツをできるだけ多く提供できるよう取り組んでいます。しばらくお待ちください。"},"ko-kr":{"text":"본 페이지 컨텐츠는 현재 선택된 언어로는 볼 수 없습니다. 가능한 빨리 현지화된 컨텐츠를 제공해 드리기 위해 노력하고 있습니다. 기다려 주셔서 감사합니다."},"zh-cn":{"text":"本页面内容目前不支持所选语言。我们将尽可能提供更多的本地化内容。敬请期待。"}}},{"_id":"616eeecda9b8227a40aa618c","key":"DTRolloutComponent","props":{"targetAudience":"100"},"created_at":"2021-10-19T16:14:05.400Z","updated_at":"2024-10-21T20:11:05.326Z"},{"_id":"653956df6e40c7d11245d051","key":"PencilBanner","props":{"pillText":"Event","disabled":false,"bannerTheme":0,"theme":"forestGreen","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner"},"created_at":"2023-10-20T17:42:11.857Z","updated_at":"2024-10-21T20:11:05.334Z","translations":{"en-us":{"theme":"forestGreen","pillText":"Event","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","bannerTheme":0,"disabled":false},"pt-br":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"es":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"it-it":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"de-de":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"fr-fr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ja-jp":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ko-kr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"zh-cn":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"}}}],"locale":"en","saved_by":{"_id":"643eae09bb4685001287c816","user_name":"kutpudeen.rahiman","permissions":{"roles":["Content Lead","Translation","MOPS Lead","admin"],"node_types":[{"type":"blog_post","actions":["translate"]},{"type":"content_block","actions":["translate"]},{"type":"digital_transformation","actions":["translate"]},{"type":"event","actions":["translate"]},{"type":"webinar","actions":["translate"]},{"type":"presentation","actions":["translate"]},{"type":"online_collateral","actions":["translate"]}],"documents":[],"collections":[]}},"tag_ids":["60cb6791cad1730d6d6f39c4"],"updateHistory":[{"time":"2024-07-05T21:00:37.029Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-05T21:05:12.515Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:11:58.751Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:30:19.899Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:31:54.653Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:43:44.420Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:44:06.997Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:45:21.072Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-12T20:52:28.916Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-12T20:53:26.641Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-08-07T13:17:35.462Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-08-08T16:19:17.364Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-08-09T09:14:14.109Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-11-24T12:58:55.262Z","user_name":"kutpudeen.rahiman","versionMessage":"WEBSITE-18050 Update Image"}],"versionMessage":"WEBSITE-18050 Update Image","authors":[],"parent_document":[],"context":{"isMobile":false,"i18n":{"First Name":"First Name","Last Name":"Last Name","Company":"Company","Country":"Country","Business Email":"Business Email","Email Address":"Email Address","Postal Code":"Postal Code","City":"City","Which best matches your current role?":"Which best matches your current role?","IT Executive (CIO, CTO, VP Engineering, etc.)":"IT Executive (CIO, CTO, VP Engineering, etc.)","Business Executive (CEO, COO, CMO, etc.)":"Business Executive (CEO, COO, CMO, etc.)","Architect":"Architect","Business Development / Alliance Manager":"Business Development / Alliance Manager","DBA":"DBA","Technical Operations":"TechOps","Director / Development Manager":"Director / Development Manager","Product / Project Manager":"Product / Project Manager","Software Developer / Engineer":"Software Developer / Engineer","Mobile Developer":"Mobile Developer","Business Analyst":"Business Analyst","Data Scientist":"Data Scientist","Student":"Student","Other":"Other","How did you hear about MongoDB.live?":"How did you hear about MongoDB.live?","Email":"Email","Facebook/Instagram":"Facebook / Instagram","Friend/Colleague":"Friend / Colleague","LinkedIn":"LinkedIn","MongoDB Sales Rep":"MongoDB Sales Rep","Other MongoDB Employee":"Other MongoDB Employee","MongoDB Website":"MongoDB Website","Phone":"Phone","Job Function":"Job Function","How are you using MongoDB?":"How are you using MongoDB?","I'm learning MongoDB":"I'm learning MongoDB","I'm building a new app":"I'm building a new app","I'm migrating an app to Atlas (Cloud)":"I'm migrating an app to Atlas (Cloud)","MongoDB Engineering Blog":"MongoDB Engineering Blog","Press Room":"Press Room","Legal Notices":"Legal Notices","Code of Conduct":"Code of Conduct","Who uses MongoDB":"Who uses MongoDB","NoSQL Database Explained":"NoSQL Database Explained","Tools":"Tools","Partners":"Partners","Customers":"Customers","About":"About","Sign In":"Sign In","Mongo, MongoDB, and the MongoDB leaf logo are registered trademarks of MongoDB, Inc.":"Mongo, MongoDB, and the MongoDB leaf logo are registered trademarks of MongoDB, Inc.","Leadership":"Leadership","2020 MongoDB, Inc. - Mongo, MongoDB, and the MongoDB leaf logo are registered trademarks of MongoDB, Inc.":"2020 MongoDB, Inc. - Mongo, MongoDB, and the MongoDB leaf logo are registered trademarks of MongoDB, Inc.","Developer Tools":"Developer Tools","Try Free":"Try Free","Education & Support":"Education & Support","Native visualization for MongoDB data":"Native visualization for MongoDB data","Server":"Server","Resources":"Resources","The database":"The database","Solutions":"Solutions","Fully managed cloud database":"Fully managed cloud database","Contact":"Contact","Find or become a partner":"Find or become a partner","MongoDB on AWS":"MongoDB on AWS","Community":"Community","A free and open document database":"A free and open document database","What is a Cloud Database?":"What is a Cloud Database?","Atlas, Realm, and more":"Atlas, Realm, and more","Investors":"Investors","What is MongoDB?":"What is MongoDB?","Use Cases":"Use Cases","Compass, Charts, Connectors, and more":"Compass, Charts, Connectors, and more","Developer Hub":"Developer Hub","Run MongoDB on Multiple Clouds with MongoDB Atlas":"Run MongoDB on Multiple Clouds with MongoDB Atlas","MongoDB Architecture Guide":"MongoDB Architecture Guide","Language APIs":"Language APIs","Popular Topics":"Popular Topics","Worldwide community events":"Worldwide community events","Follow Us":"Follow Us","The MongoDB Community discussion forums":"The MongoDB Community discussion forums","Free online courses from MongoDB":"Free online courses from MongoDB","MongoDB on Google Cloud":"MongoDB on Google Cloud","Application Development Services":"Application Development Services","FAQ":"FAQ","Learn":"Learn","Cloud-native full-text search engine":"Cloud-native full-text search engine","Developer best practices, trends, insights":"Developer best practices, trends, insights","Advanced features and security":"Advanced features and security","Training":"Training","Building a REST API with MongoDB Realm":"Building a REST API with MongoDB Realm","Easy integrations to your data estate":"Easy integrations to your data estate","Connect, configure and work with MongoDB":"Connect, configure and work with MongoDB","Connectors":"Connectors","Community Server":"Community Server","Query and combine AWS S3 and MongoDB data":"Query and combine AWS S3 and MongoDB data","University":"University","Trust Center":"Trust Center","Careers":"Careers","Events":"Events","How MongoDB is used":"How MongoDB is used","Enterprise Server":"Enterprise Server","Installation":"Installation","How to Guides":"How to Guides","Support":"Support","Consulting":"Consulting","Office Locations":"Office Locations","Start here":"Start here","Webinars, white papers, datasheets, and more":"Webinars, white papers, datasheets, and more","On-prem management platform for MongoDB":"On-prem management platform for MongoDB","Updates, tutorials, people":"Updates, tutorials, people","View Course Catalog":"View Course Catalog","MongoDB Manual":"MongoDB Manual","Pricing":"Pricing","Instructor-led sessions on your schedule":"Instructor-led sessions on your schedule","Get started in minutes":"Get started in minutes","Accelerate success with MongoDB":"Accelerate success with MongoDB","Migrate to MongoDB Atlas":"Migrate to MongoDB Atlas","Docs":"Docs","Charts":"Charts","Privacy Notice":"Privacy Notice","Certification":"Certification","GUI for MongoDB":"GUI for MongoDB","Security Information":"Security Information","Learn more":"Learn more","terms of service":"terms of service","privacy policy":"privacy policy","Sign in":"Sign in","or":"or","Your Work Email":"Your Work Email","to login":"to login","I agree to the":"I agree to the","and":"and","Click here":"Click here","Already have an account":"Already have an account","Sign up with Google":"Sign up with Google","Sign up with Github":"Sign up with Github","Your Company (optional)":"Your Company (optional)","8 characters minimum":"8 characters minimum","Sitemap":"Sitemap","Read White Paper":"Read White Paper","View Datasheet":"View Datasheet","Read Analyst Report":"Read Analyst Report","Email Me the PDF":"Email Me the PDF","More like this":"More like this","View all resources":"View all resources","language":"language","share this":"share this","Access White Paper":"Access White Paper","Access Datasheet":"Access Datasheet","Access Analyst Report":"Access Analyst Report"},"promoBanner":{"_id":"601c7536f53e6b3af09679d3","key":"PromoBanner","created_at":"2021-02-04T22:29:10.420Z","props":{"type":4,"typeColor":0,"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","titleColor":0,"background":1,"disabled":false,"eventBranded":false,"eventBrandedButtonImg":0},"updated_at":"2024-10-21T20:11:05.312Z","translations":{"en-us":{"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","eventBranded":false,"disabled":false,"type":4,"eventBrandedButtonImg":0,"typeColor":0,"background":1,"titleColor":0},"pt-br":{"title":"Register for MongoDB.live today!","type":0,"titleColor":0},"es":{"title":"Register for MongoDB.live today!"},"it-it":{"title":"Register for MongoDB.live today!"},"de-de":{"title":"Register for MongoDB.live today!"},"fr-fr":{"title":"Register for MongoDB.live today!"},"ja-jp":{"title":"Register for MongoDB.live today!"},"ko-kr":{"title":"Register for MongoDB.live today!"},"zh-cn":{"title":"Register for MongoDB.live today!"}}},"pencilBanner":{"_id":"653956df6e40c7d11245d051","key":"PencilBanner","props":{"pillText":"Event","disabled":false,"bannerTheme":0,"theme":"forestGreen","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner"},"created_at":"2023-10-20T17:42:11.857Z","updated_at":"2024-10-21T20:11:05.334Z","translations":{"en-us":{"theme":"forestGreen","pillText":"Event","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","bannerTheme":0,"disabled":false},"pt-br":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"es":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"it-it":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"de-de":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"fr-fr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ja-jp":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ko-kr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"zh-cn":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"}}},"locale":"en","urlLocale":"en","url":"/resources/basics/vector-embeddings","cookies":{},"translationFallbackBanner":{"_id":"60c127b5527761a42edca7bb","key":"TranslationFallbackBanner","created_at":"2021-06-09T20:42:29.953Z","updated_at":"2024-10-21T20:11:05.319Z","props":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"translations":{"en-us":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"pt-br":{"text":"O conteúdo desta página não está disponível no idioma selecionado no momento. No entanto, estamos comprometidos em oferecer o máximo de conteúdo localizado possível. Agradecemos a paciência."},"es":{"text":"El contenido de esta página no está disponible actualmente en el idioma seleccionado. Sin embargo, nos comprometemos a proporcionar la mayor cantidad de contenido localizado posible. Gracias por tu paciencia."},"it-it":{"text":"I contenuti di questa pagina non sono attualmente disponibili nella lingua selezionata. Tuttavia, ci impegniamo a fornire il maggior numero possibile di contenuti localizzati. Grazie per la pazienza."},"de-de":{"text":"Die Inhalte dieser Seite sind derzeit nicht in der gewählten Sprache verfügbar. Wir arbeiten jedoch daran, so viele lokalisierte Inhalte wie möglich bereitzustellen. Vielen Dank für Ihre Geduld."},"fr-fr":{"text":"Le contenu de cette page n'est actuellement pas disponible dans la langue sélectionnée. Nous mettons toutefois tout en œuvre pour proposer autant de contenu localisé que possible. Merci de votre patience."},"ja-jp":{"text":"現在、このページの選択した言語のコンテンツはありません。ローカライズされたコンテンツをできるだけ多く提供できるよう取り組んでいます。しばらくお待ちください。"},"ko-kr":{"text":"본 페이지 컨텐츠는 현재 선택된 언어로는 볼 수 없습니다. 가능한 빨리 현지화된 컨텐츠를 제공해 드리기 위해 노력하고 있습니다. 기다려 주셔서 감사합니다."},"zh-cn":{"text":"本页面内容目前不支持所选语言。我们将尽可能提供更多的本地化内容。敬请期待。"}}},"ip":"8.222.208.146"},"hideMenu":false,"querystring":{},"localizedDocuments":[{"_id":"66885d2e687f6d6235b699fd","url":"resources/basics/vector-embeddings","components":[{"key":"Nav","props":{"left":[{"title":"Cloud","links":[{"title":"Atlas","text":"Fully managed cloud database","href":"/cloud/atlas"},{"title":"Atlas Data Lake","text":"Query and combine AWS S3 and MongoDB data","href":"/atlas/data-lake"},{"title":"Atlas Search","text":"Cloud-native full-text search engine","href":"/atlas/search"},{"title":"Realm","text":"Application Development Services","href":"/realm"},{"title":"Charts","text":"Native visualization for MongoDB data","href":"/products/charts"},{"title":"Atlas for Government","text":"Atlas for US Government workloads","href":"/cloud/atlas/government"}]},{"title":"Software","links":[{"title":"Community Server","text":"A free and open document database","href":"/try/download/community"},{"title":"Enterprise Server","text":"Advanced features and security","href":"/try/download/enterprise"},{"title":"Developer Tools","text":"Connect, configure and work with MongoDB","href":"/developer-tools"},{"title":"Compass","text":"GUI for MongoDB","href":"/products/compass"},{"title":"Ops Manager","text":"On-prem management platform for MongoDB","href":"/products/ops-manager"},{"title":"Connectors","text":"Easy integrations to your data estate","href":"/connectors"}]},{"title":"Pricing","links":[],"href":"/pricing"},{"title":"Learn","links":[{"title":"What is MongoDB?","text":"Start here","href":"/what-is-mongodb"},{"title":"University","href":"https://university.mongodb.com","text":"Free online courses from MongoDB"},{"title":"Blog","href":"/blog","text":"Updates, tutorials, people"},{"title":"Developer Hub","href":"https://developer.mongodb.com","text":"Developer best practices, trends, insights"},{"title":"Resources","href":"/resources","text":"Webinars, white papers, datasheets, and more"},{"title":"Training","href":"/training","text":"Instructor-led sessions on your schedule"},{"title":"Events","href":"/events","text":"Worldwide community events"},{"title":"Community","href":"https://community.mongodb.com","text":"The MongoDB Community discussion forums"}]},{"title":"Solutions","links":[{"title":"Customers","text":"Who uses MongoDB","href":"/who-uses-mongodb"},{"title":"Use Cases","text":"How MongoDB is used","href":"/use-cases"},{"title":"Consulting","text":"Accelerate success with MongoDB","href":"/products/consulting"},{"title":"Partners","text":"Find or become a partner","href":"/partners"}]},{"title":"Docs","links":[{"title":"Cloud","text":"Atlas, Realm, and more","href":"https://docs.mongodb.com/cloud/"},{"title":"Server","href":"https://docs.mongodb.com/manual/","text":"The database"},{"title":"Drivers","text":"Language APIs","href":"https://docs.mongodb.com/ecosystem/drivers/"},{"title":"Tools","text":"Compass, Charts, Connectors, and more","href":"https://docs.mongodb.com/tools/"},{"title":"How to Guides","text":"Get started in minutes","href":"https://docs.mongodb.com/guides/"}]}],"right":[{"title":"Contact","href":"/contact","button":false},{"title":"Sign In","href":"https://cloud.mongodb.com/user","button":false},{"title":"Try Free","href":"/try","button":true}],"mobile":[{"text":"Contact","href":"/contact"}],"navType":"","notSticky":false,"banner":{"bannerTheme":"","pillText":"","bannerText":"","href":""}},"id":"f93b19ee-9a15-44b6-9563-f309b7d880f0"},{"key":"Header","props":{"h1":"What are Vector Embeddings? ","titleColor":0,"content":"","cta":{"href":"https://www.mongodb.com/cloud/atlas/register","text":"Get started free","openInNewWindow":true,"faux":false},"cta2":{"href":"","text":"","openInNewWindow":false,"faux":false},"imgsrc":"https://webassets.mongodb.com/_com_assets/cms/header-au039oiqsr.svg","bgColor":"","bgImage":""},"id":1720212972605},{"key":"Article","props":{"content":"Vector embeddings are mathematical representations of text created by translating words or sentences into numbers — a language that computers can understand. They bridge the rich, nuanced world of human language (text, images, speeches, videos etc.) and the precise environment of machine learning models (numbers) by representing data points.\n\n![An image of vector embeddings including unstructured data and encoder.](https://webimages.mongodb.com/_com_assets/cms/m3n5dd9oq8zrgpot7-vector-embedding.png?auto=format%252Ccompress)\n\n<br>\n\nMost often used in natural language processing (NLP), vector embeddings allow machine learning algorithms to analyze information much like humans but at a scale and speed far beyond our capabilities. \nAlthough they also work with images, audio processing, bioinformatics, and recommendation systems, this article will focus on word vector embeddings in natural language processing using a machine learning model. \n\n<br>\n\n**Table of contents**\n\n- [What is natural language processing?](#what-is-natural-language-processing)\n- [Types of vector embeddings in NLP](#types-of-vector-embeddings-in-nlp)\n- [What is a vector?](#what-is-a-vector-)\n- [Vector embeddings in multidimensional spaces](#vector-embeddings-in-multidimensional-spaces)\n- [Dimensionality in vector embeddings](#dimensionality-in-vector-embeddings)\n- [Advantages of vector embeddings in real-world scenarios](#advantages-of-vector-embeddings-in-realworld-scenarios)\n- [Challenges and limitations](#challenges-and-limitations)\n- [Using vector embeddings in other applications](#using-vector-embeddings-in-other-applications-)\n- [Why use MongoDB Atlas Vector Search for vector similarity search?](#why-use-mongodb-atlas-vector-search-for-vector-similarity-search)\n- [Conclusion](#conclusion)\n- [FAQs](#faqs)\n\n","textalign":"","background":""},"id":1720212947522},{"key":"Article","props":{"content":"##What is natural language processing?\nNLP is a type of artificial intelligence that uses vector embeddings in conjunction with machine learning algorithms to evaluate, understand, and interpret human language. This combination achieves comprehension and interaction that mirrors human ability but at a scale and speed far beyond our capabilities. NLP excels in tasks such as interpreting text from social media, translating languages, and powering conversational agents.\n","textalign":"","background":""},"id":1720213124707},{"key":"Article","props":{"content":"##Types of vector embeddings in NLP\nBelow are a few examples of the diverse vector embedding techniques instrumental in advancing NLP, each bringing its strengths to various language understanding challenges.\n\n- **Word2Vec**: Developed by Google, Word2Vec captures the context of words within documents. It’s beneficial for tasks that require understanding word associations and meanings based on their usage in sentences.\n\n- **GloVe (global vectors for word representation)**: GloVe is unique in its approach as it analyzes word co-occurrences over the whole corpus for training, enabling it to capture global statistics of words. It’s particularly useful for tasks that involve semantic similarity between words.\n\n- **BERT (Bidirectional Encoder Representations from Transformers)**: Developed by Google, BERT represents a breakthrough in contextually aware embeddings. It looks at the context from both sides of a word in a sentence, making it highly effective for sophisticated tasks like sentiment analysis and question answering.","textalign":"","background":""},"id":1720213145679},{"key":"Article","props":{"content":"##What is a vector? \nTo truly appreciate what vector embeddings are and how they work, it's essential first to understand what a vector is in this setting. Think of a vector as a point in space with direction and magnitude. It's like a dot on a map with specific coordinates. These data points aren't just random numbers; they represent different characteristics or features of the data types that the vector represents. \n","textalign":"","background":""},"id":1720213160578},{"key":"Article","props":{"content":"##Vector embeddings in multidimensional spaces\nNow that the basics of vectors and data points have been introduced, it's important to note that in the context of vector representations for text, more than one vector embedding or one dimension is used. When text — words, phrases, or entire documents — are converted into vectors, each piece of text plots as a point in a vast, multidimensional space. This space isn't like a typical three-dimensional space we're familiar with; it has many more dimensions, each representing a different aspect of the text's meaning or usage. \n\nImagine a map where words with similar meanings or usages are mapped closer together, making it easier to see their relationships or similar data points.\n\nTurning text into vector embeddings is a game-changer for machine learning algorithms. These algorithms are great at dealing with numbers — they can spot patterns, make comparisons, and draw conclusions from numerical data.\n","textalign":"","background":""},"id":1720213183671},{"key":"Article","props":{"content":"##Dimensionality in vector embeddings\nLet's dig a little deeper into the concept of dimensionality, which we introduced above. Think of dimensionality in vector embeddings like the resolution of a photo. High-resolution photos are more detailed and precise, but they take up more space on your phone and require more processing power. Similarly, in vector embeddings, more dimensions mean that the representation of words or phrases can capture more details and nuances of language.\n\n<br>\n\n###High-dimensional embeddings\nHigh-dimensional embeddings are like high-resolution photos. They have hundreds or even thousands of dimensions, allowing them to capture much information about a word or phrase. Each dimension can represent a different aspect of a word's meaning or use. This detailed representation is excellent for complex tasks in natural language processing, where understanding subtle differences in language is crucial. \n\nHowever, like high-resolution photos, these embeddings require more computer memory and processing power. Also, there's a risk of \"overfitting\" — think of it like a camera that focuses on capturing every tiny detail and fails to recognize common, everyday objects. In machine learning, the model might get too tailored to its training data and perform poorly on new, unseen data.\n\n<br>\n\n###Low-dimensional embeddings\nOn the other hand, low-dimensional embeddings are like lower-resolution photos. They have fewer dimensions, so they use less computer memory and process more quickly, which is excellent for applications that need to run fast or have limited resources. But just like lower-resolution photos can miss finer details, these embeddings might not capture all the subtle nuances of language. Depending on the task, they provide a more general picture, which can sometimes be enough.\n\nChoosing the proper dimensionality for creating vector embeddings is a balance. It's about weighing the need for detail against the need for efficiency and the ability of the model to perform well on new, unseen data. Finding the right balance often involves trial and error and depends on the specific task and the data. It's a crucial part of developing effective NLP solutions, requiring a thoughtful approach to meet both the linguistic needs of the task and the practical limitations of technology.\n","textalign":"","background":""},"id":1720213190910},{"key":"Article","props":{"content":"##Advantages of vector embeddings in real-world scenarios\nVector embeddings have opened up a world of possibilities in how machines interact with human language. They make technology more intuitive and natural, enriching interactions across digital platforms and tools. Below are a few applications highlighting how vector embeddings are used today. \n\n<br>\n\n###Sentiment analysis\nSentiment analysis is like a digital mood ring. Businesses use it to understand how people feel about their products or services by analyzing the tone of customer reviews and social media posts. Vector embeddings help computers catch subtle emotional cues in text, distinguishing genuine praise from sarcasm, even when the words are similar.\n\n<br>\n\n###Machine translation\nVector embeddings are the backbone of translation apps. They help computers grasp the complexities and nuances of different languages. When a sentence is translated from one language to another, it's not just about swapping words; it's about conveying the same meaning, tone, and context. Vector embeddings are crucial in achieving this.\n\n<br>\n\n###Chatbots and virtual assistants\nAre you curious how virtual assistants like Siri or Alexa understand and respond to your queries so well? This functionality is primarily due to vector embeddings. They enable artificial intelligence (AI) systems to process what you're saying, figure out what you mean, and respond in a way that makes sense.\n\n<br>\n\n###Information retrieval\nThis category covers everything from search engines to recommendation systems. Vector embeddings help these systems understand what is being searched, not just by matching keywords but by grasping the context of the query. This way, the information or recommendations will more likely be relevant.\n\n<br>\n\n###Text classification\nText classification can filter emails, categorize news articles, and even tag social media posts. Vector embeddings assist in sorting text into different categories by understanding the underlying themes and topics, making it easier for algorithms to decide, for example, if an email is spam.\n\n<br>\n\n###Speech recognition\nRegarding converting spoken words into written text, vector embeddings play a crucial role. They help capture the spoken words accurately, considering how the same word can be pronounced or used in different contexts, leading to more accurate transcriptions.\n","textalign":"","background":""},"id":1720213253255},{"key":"Article","props":{"content":"##Challenges and limitations\nWhile vector embeddings are a powerful tool in NLP, they are not without their challenges. Addressing these issues is crucial for ensuring that these technologies are effective, fair, and up-to-date, requiring continuous effort and innovation in the field. Let's explore these challenges and limitations, especially in how vector embeddings interact with and process human language.\n\n<br>\n\n###Handling out-of-vocabulary words\nOne of the trickiest issues with vector embeddings is dealing with words that the system has never seen before, often called \"out-of-vocabulary\" words. It's like encountering a word in a foreign language you've never learned. For a computer, these new words can be a fundamental stumbling block. The system might struggle to understand and place them correctly in the context of what it already knows.\n\n<br>\n\n###Bias in embeddings\nLike humans, computers can be biased, especially when they learn from data reflecting human prejudices. When vector embeddings train on text data from the internet or other human-generated sources, they can inadvertently pick up and even amplify these biases. This possibility is significant because it can lead to unfair or stereotypical representations in various applications, like search engines or AI assistants.\n\n<br>\n\n###Complexity of maintaining and updating models\nKeeping vector embedding models up-to-date and relevant is no small task. Language is constantly evolving — new words pop up, old words fade away, and meanings change. Ensuring these models stay current is like updating a constantly evolving dictionary. It requires ongoing work and resources, making it a complex and challenging aspect of working with vector embeddings.\n\n<br>\n\n###Contextual ambiguity\nWhile vector embeddings are good at capturing meaning, they sometimes struggle with words with multiple meanings based on context. For instance, the word \"bat\" can refer to an animal or sports equipment, and without sufficient context, the model might not accurately capture the intended use.\n\n<br>\n\n###Resource intensity\nTraining sophisticated vector embedding models requires significant computational resources, which can be a barrier, especially for smaller organizations or individual researchers who might not have access to the necessary computing power.\n\n<br>\n\n###Data quality and availability\nThe effectiveness of vector embeddings heavily depends on the quality and quantity of the training data. The embeddings might not be as accurate or helpful in languages or domains where data is scarce or of poor quality.\n\n<br>\n\n###Transferability across languages\nVector embeddings trained in one language may not transfer well to another, especially for structurally different languages. These structural differences challenge multilingual applications or languages with limited resources.\n\n<br>\n###Model interpretability\nUnderstanding why a vector embedding model behaves a certain way or makes specific decisions can be challenging. This lack of interpretability can be a significant issue, especially in applications where understanding the model's reasoning is crucial.\n\n<br>\n\n###Scalability\nAs the amount of data and the complexity of tasks increase, scaling vector embedding models while maintaining performance and efficiency can be challenging.\n\n<br>\n\n###Dependency on training data\n\nVector embeddings can only be as good as their training data. If the training data is limited or biased, the embeddings will inherently reflect those limitations or biases.\n","textalign":"","background":""},"id":1720213273484},{"key":"Article","props":{"content":"##Using vector embeddings in other applications \nWhile vector embeddings are most prominently used in NLP, they are also used in other ways. Here are a few other areas where vector embeddings are employed. \n\n<br>\n\n###Computer vision\nIn image processing and computer vision, embeddings represent images or parts of images. Similar to how they capture the essence of words in NLP, embeddings in computer vision capture essential features of images, enabling tasks like image recognition, classification, and similarity detection.\n\n<br>\n\n###Recommendation systems\nVector embeddings also show up in recommendation systems, such as those on e-commerce or streaming platforms. They help understand user preferences and item characteristics by representing users and items in a vector space, enabling the system to make personalized recommendations based on similarity.\n\n<br>\n\n###Bioinformatics\nIn bioinformatics, embeddings can represent biological data, such as gene sequences or protein structures. These embeddings help in various predictive tasks, like understanding gene function or protein-protein interactions.\n\n<br>\n\n###Graph analysis\nIn network and graph analysis, embeddings represent nodes and edges of a graph, which is helpful in social network analysis, link prediction, and understanding the structure and dynamics of complex systems.\n\n<br>\n\n###Time series analysis\nVector embeddings work in analyzing time series data, such as financial market trends or sensor data, by capturing temporal patterns and dependencies in a vector space.\n\nThese diverse applications show that the concept of embeddings is a versatile tool in the broader field of machine learning and data science, not limited to just text and language processing.\n","textalign":"","background":""},"id":1720213303077},{"key":"Article","props":{"content":"##Why use MongoDB Atlas Vector Search for vector similarity search?\n\n**Overview**\n\nMongoDB Atlas Vector Search is an advanced tool designed to handle complex vector similarity searches. It leverages the strengths of MongoDB's flexible data model and robust indexing capabilities, making it a powerful solution for various search and generative AI applications requiring vector search.\n\n**Key benefits**\n\n1. Seamless integration with MongoDB: Atlas Vector Search is built into MongoDB, allowing you to use the same database for both structured and unstructured data. This integration simplifies your architecture and data management processes.\n\n2. Scalability: MongoDB Atlas provides a highly scalable environment that can handle large volumes of data, making it ideal for applications requiring extensive vector searches.\n\n2. Flexible indexing: MongoDB's indexing capabilities enable efficient storage and retrieval of vector data, ensuring fast and accurate search results.\n\n4. Multi-cloud availability: Atlas Vector Search is available across major cloud providers, ensuring flexibility and reliability.\n\n5. Security: Benefit from MongoDB's advanced security features, including encryption at rest and in transit, role-based access control, and comprehensive auditing.\n\n<br>\n\n###Similarity algorithms supported\nCosine similarity: This measures the cosine of the angle between two vectors. It is particularly useful for comparing documents in text analysis, as it considers the orientation rather than the magnitude of the vectors.\n\nEuclidean distance: This calculates the straight-line distance between two points in a multidimensional space. It is a simple and intuitive measure of similarity, often used in clustering and classification tasks.\n\nDot product: This computes the sum of the products of the corresponding entries of two sequences of numbers. It is used in various applications, including machine learning and recommendation systems, to measure the similarity between two vectors.\n","textalign":"","background":""},"id":1720213320309},{"key":"Article","props":{"content":"##Conclusion\nVector embeddings represent a significant leap in how machines process and understand human language and other complex data types. From enhancing the capabilities of NLP in understanding text to their applications in fields like computer vision and bioinformatics, vector embeddings have proven to be invaluable tools. As technology evolves, so will the sophistication and utility of vector embeddings.\n\nBy storing vector embeddings in documents alongside metadata and contextual app data in a single, unified, fully managed, secure platform, developers can enjoy a seamless, flexible, and simplified experience. MongoDB’s robust integrations with all major AI services and cloud providers allow developers to use the embedding model of their choice and then perform indexing and searching, building apps efficiently and securely all in one place. This streamlined approach empowers developers to avoid the complexity of dealing with multiple platforms and focus more on building effective search and generative AI applications for their organizations. See how [MongoDB Vector Search](/products/platform/atlas-vector-search) works, and visit the Atlas Vector Search [Quick Start guide](https://www.mongodb.com/docs/atlas/atlas-vector-search/tutorials/vector-search-quick-start/?tck=ai_as_web) to create your first index in minutes.","textalign":"","background":""},"id":1720213343098},{"key":"Article","props":{"content":"##FAQs","textalign":"","background":""},"id":1720213371255},{"key":"Details","props":{"items":[{"titleColor":0,"title":"What is semantic search?","text":"\n[Semantic search](https://www.mongodb.com/resources/basics/semantic-search) refers to a search technique that goes beyond keyword matching to understand the intent and contextual meaning of the search query. Instead of just looking for exact word matches, semantic search considers factors like the context of words in the query, the relationship between words, synonyms, and the overall meaning behind the query. This approach allows for more accurate and relevant search results, as it aligns more closely with how humans understand and use language."},{"titleColor":0,"title":"Does a reverse image search involve vector embeddings?","text":"\nYes, in a reverse image search, images are transformed into vector embeddings, which are used to compare and find similar images in a database, making the search process efficient and accurate."},{"titleColor":0,"title":"What is anomaly detection?","text":"\nAnomaly detection is a technique used in data analysis and various applications to identify patterns that do not conform to expected behavior. These nonconforming patterns are often referred to as anomalies, outliers, or exceptions."}],"bgColor":"","bgImage":""},"id":1720213389786},{"key":"Footer","props":{"toggle":0,"column1":{"title":"Resources","maxWidth":"185","hasIcons":0,"className":"","items":[{"href":"/nosql-explained","text":"NoSQL Database Explained","isTarget":""},{"href":"/collateral/mongodb-architecture-guide","text":"MongoDB Architecture Guide","isTarget":""},{"href":"/products/mongodb-enterprise-advanced","text":"MongoDB Enterprise Advanced","isTarget":""},{"href":"/cloud/atlas","text":"MongoDB Atlas","isTarget":""},{"href":"/cloud/stitch","text":"MongoDB Stitch","isTarget":""},{"href":"//engineering.mongodb.com/","text":"MongoDB Engineering Blog","isTarget":"true"}]},"column2":{"title":"Education & Support","maxWidth":"150","hasIcons":0,"className":"","items":[{"href":"//university.mongodb.com/courses/catalog","text":"View Course Catalog","isTarget":"true"},{"href":"//university.mongodb.com/certification","text":"Certification","isTarget":"true"},{"href":"//docs.mongodb.com/manual/","text":"MongoDB Manual","isTarget":"true"},{"href":"//docs.mongodb.com/manual/installation/","text":"Installation","isTarget":"true"},{"href":"//support.mongodb.com/welcome","text":"Support","isTarget":""},{"href":"/faq","text":"FAQ","isTarget":""}]},"column3":{"title":"Popular Topics","maxWidth":"300","hasIcons":0,"className":"be-ix-link-block","items":[{"href":"/cloud/atlas/aws-mongodb","text":"MongoDB on AWS","isTarget":""},{"href":"/cloud/atlas/mongodb-google-cloud","text":"MongoDB on Google Cloud","isTarget":""},{"href":"/cloud/atlas/multicloud-data-distribution","text":"Run MongoDB on Multiple Clouds with MongoDB Atlas","isTarget":""},{"href":"/cloud/atlas/migrate","text":"Migrate to MongoDB Atlas","isTarget":""},{"href":"/cloud-database","text":"What is a Cloud Database?","isTarget":""},{"href":"/blog/post/building-a-rest-api-with-mongodb-stitch","text":"Building a REST API with MongoDB Stitch","isTarget":""}]},"column4":{"title":"About","maxWidth":"100","hasIcons":0,"className":"","items":[{"href":"/company","text":"MongoDB, Inc.","isTarget":""},{"href":"/leadership","text":"Leadership","isTarget":""},{"href":"/company/newsroom","text":"Press Room","isTarget":""},{"href":"/careers","text":"Careers","isTarget":""},{"href":"https://investors.mongodb.com","text":"Investors","isTarget":""},{"href":"/legal/legal-notices","text":"Legal Notices","isTarget":""},{"href":"/legal/privacy-policy","text":"Privacy Notice","isTarget":""},{"href":"/security","text":"Security Information","isTarget":""},{"href":"/cloud/trust","text":"Trust Center","isTarget":""},{"href":"/office-locations","text":"Office Locations","isTarget":""},{"href":"/community-code-of-conduct","text":"Code of Conduct","isTarget":""}]},"column5":{"title":"Follow Us","maxWidth":"120","hasIcons":1,"className":"","items":[{"href":"//facebook.com/mongodb","text":"Facebook","isTarget":"true"},{"href":"//github.com/mongodb","text":"Github","isTarget":"true"},{"href":"//youtube.com/user/mongodb","text":"Youtube","isTarget":"true"},{"href":"//twitter.com/mongodb","text":"Twitter","isTarget":"true"},{"href":"//www.linkedin.com/company/mongodbinc/","text":"LinkedIn","isTarget":"true"},{"href":"//slackpass.io/mongo-db","text":"Slack","isTarget":"true"},{"href":"//stackoverflow.com/tags/mongodb/info","text":"StackOverflow","isTarget":"true"}]}},"id":"aecfeaf5-65f6-4623-ad36-94949d7aef26"}],"created_at":"2024-07-05T20:53:02.596Z","meta":{"title":"What Are Vector Embeddings? | MongoDB","title#localised":true,"description":"Learn the basics of vector embeddings, its role in AI, and how MongoDB utilizes this technology.","description#localised":true},"node_type":"content_block","owners":[],"published_at":"2024-07-08T16:45:21.072Z","status":"published","updated_at":"2024-11-24T12:58:55.262Z","cms":{"editedURL":true},"draft":true,"globals":[{"_id":"6001f22ac1f95e773a0e0044","key":"AccountLogin","created_at":"2021-01-15T19:51:06.717Z","props":{"title":"MongoDB Stands with the Black Community, changes","subtitle":"Join MongoDB in supporting organizations that are fighting for racial justice and equal opportunity","cta":{"text":"Join Now","href":"https://mongodbforjustice.mongodb.events/","openInNewWindow":false,"faux":false},"image":{"desktop":"https://account.mongodb.com/static/images/auth/racial_justice_desktop_login.png","mobile":"https://account.mongodb.com/static/images/auth/racial_justice_mobile.png"},"artist":"Artwork by [Lo Harris](http://loharris.com/)"},"updated_at":"2024-10-21T20:11:05.302Z"},{"_id":"601c7536f53e6b3af09679d3","key":"PromoBanner","created_at":"2021-02-04T22:29:10.420Z","props":{"type":4,"typeColor":0,"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","titleColor":0,"background":1,"disabled":false,"eventBranded":false,"eventBrandedButtonImg":0},"updated_at":"2024-10-21T20:11:05.312Z","translations":{"en-us":{"title":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. Learn more >>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","eventBranded":false,"disabled":false,"type":4,"eventBrandedButtonImg":0,"typeColor":0,"background":1,"titleColor":0},"pt-br":{"title":"Register for MongoDB.live today!","type":0,"titleColor":0},"es":{"title":"Register for MongoDB.live today!"},"it-it":{"title":"Register for MongoDB.live today!"},"de-de":{"title":"Register for MongoDB.live today!"},"fr-fr":{"title":"Register for MongoDB.live today!"},"ja-jp":{"title":"Register for MongoDB.live today!"},"ko-kr":{"title":"Register for MongoDB.live today!"},"zh-cn":{"title":"Register for MongoDB.live today!"}}},{"_id":"60c127b5527761a42edca7bb","key":"TranslationFallbackBanner","created_at":"2021-06-09T20:42:29.953Z","updated_at":"2024-10-21T20:11:05.319Z","props":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"translations":{"en-us":{"text":"The contents of this page are not currently available in the selected language. However, we are committed to providing as much localized content as possible. Thanks for your patience."},"pt-br":{"text":"O conteúdo desta página não está disponível no idioma selecionado no momento. No entanto, estamos comprometidos em oferecer o máximo de conteúdo localizado possível. Agradecemos a paciência."},"es":{"text":"El contenido de esta página no está disponible actualmente en el idioma seleccionado. Sin embargo, nos comprometemos a proporcionar la mayor cantidad de contenido localizado posible. Gracias por tu paciencia."},"it-it":{"text":"I contenuti di questa pagina non sono attualmente disponibili nella lingua selezionata. Tuttavia, ci impegniamo a fornire il maggior numero possibile di contenuti localizzati. Grazie per la pazienza."},"de-de":{"text":"Die Inhalte dieser Seite sind derzeit nicht in der gewählten Sprache verfügbar. Wir arbeiten jedoch daran, so viele lokalisierte Inhalte wie möglich bereitzustellen. Vielen Dank für Ihre Geduld."},"fr-fr":{"text":"Le contenu de cette page n'est actuellement pas disponible dans la langue sélectionnée. Nous mettons toutefois tout en œuvre pour proposer autant de contenu localisé que possible. Merci de votre patience."},"ja-jp":{"text":"現在、このページの選択した言語のコンテンツはありません。ローカライズされたコンテンツをできるだけ多く提供できるよう取り組んでいます。しばらくお待ちください。"},"ko-kr":{"text":"본 페이지 컨텐츠는 현재 선택된 언어로는 볼 수 없습니다. 가능한 빨리 현지화된 컨텐츠를 제공해 드리기 위해 노력하고 있습니다. 기다려 주셔서 감사합니다."},"zh-cn":{"text":"本页面内容目前不支持所选语言。我们将尽可能提供更多的本地化内容。敬请期待。"}}},{"_id":"616eeecda9b8227a40aa618c","key":"DTRolloutComponent","props":{"targetAudience":"100"},"created_at":"2021-10-19T16:14:05.400Z","updated_at":"2024-10-21T20:11:05.326Z"},{"_id":"653956df6e40c7d11245d051","key":"PencilBanner","props":{"pillText":"Event","disabled":false,"bannerTheme":0,"theme":"forestGreen","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner"},"created_at":"2023-10-20T17:42:11.857Z","updated_at":"2024-10-21T20:11:05.334Z","translations":{"en-us":{"theme":"forestGreen","pillText":"Event","bannerText":"Join us at AWS re:Invent 2024! Learn how to use MongoDB for AI use cases. <mark>Learn more >>\u003C/mark>","href":"https://www.mongodb.com/events/aws-reinvent?tck=pencil_banner","bannerTheme":0,"disabled":false},"pt-br":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"es":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"it-it":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"de-de":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"fr-fr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ja-jp":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"ko-kr":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"},"zh-cn":{"bannerText":"MongoDB World is back in NYC June 7 - 9!"}}}],"locale":"en","saved_by":{"_id":"643eae09bb4685001287c816","user_name":"kutpudeen.rahiman","permissions":{"roles":["Content Lead","Translation","MOPS Lead","admin"],"node_types":[{"type":"blog_post","actions":["translate"]},{"type":"content_block","actions":["translate"]},{"type":"digital_transformation","actions":["translate"]},{"type":"event","actions":["translate"]},{"type":"webinar","actions":["translate"]},{"type":"presentation","actions":["translate"]},{"type":"online_collateral","actions":["translate"]}],"documents":[],"collections":[]}},"tag_ids":["60cb6791cad1730d6d6f39c4"],"updateHistory":[{"time":"2024-07-05T21:00:37.029Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-05T21:05:12.515Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:11:58.751Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:30:19.899Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:31:54.653Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:43:44.420Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:44:06.997Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-08T16:45:21.072Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-12T20:52:28.916Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-07-12T20:53:26.641Z","user_name":"eric.gamble","versionMessage":"Embeddings"},{"time":"2024-08-07T13:17:35.462Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-08-08T16:19:17.364Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-08-09T09:14:14.109Z","user_name":"sergii.chubatiuk","versionMessage":"WEBSITE-17446 update"},{"time":"2024-11-24T12:58:55.262Z","user_name":"kutpudeen.rahiman","versionMessage":"WEBSITE-18050 Update Image"}],"versionMessage":"WEBSITE-18050 Update Image"}]}</script> <script> var script = document.createElement('script'); script.src = "https://d2c7xlmseob604.cloudfront.net/tracker.min.js"; script.onload = function() { SmartlingContextTracker.init({ orgId: '7v+wvebTu7n8KdlwsG2ojQ', }); } document.body.append(script); </script> </html>