Embeddings
Turn text (and sometimes images) into vectors so you can search, cluster and retrieve by meaning.
Trends
Gemini API gained the most ground in package downloads: 4.9× against the median npm package over a year.
Public counts each month: pull requests on GitHub, package downloads, stars, Product Hunt followers. They show direction, not market share.
- Oct 8 · Action needed Gemini API shuts down the deep-research-pro-preview-12-2025 agent on October 23 and now routes gemini-3.7-flash and gemini-3.5-flash to their successors; update the agent and model strings.
- Oct 6 · Action needed Gemini API deprecates the gemini-3.1-flash-image model, with no shutdown date yet; move image generation to gemini-nano-banana-2.1.
- Oct 1 · Action needed OpenAI removes gpt-5.1, gpt-5.3-codex and gpt-5.4-nano from the API on April 1, 2027; move to gpt-6-sol or gpt-6-luna.
- Oct 1 · Action needed OpenAI removes the tts-1, tts-1-hd and gpt-4o-mini-tts text-to-speech models on January 6, 2027; move to gpt-realtime-2.1-mini.
The real choice
Stay with the vendor you already use, pick a retrieval specialist, or run an open model yourself. Using your main LLM provider's embeddings means one key and one bill. Specialists offer models tuned for retrieval, many languages or specific domains, plus rerankers, at the cost of one more vendor. Open embedding models run on your own hardware for free and keep text private, but you host and scale them.
Pick by situation
Tap the ones that are you — the tools that fit light up below.- You already use OpenAI or Gemini for text and want no new vendor
OpenAI API
Gemini API - Your users search in many languages, or you want embeddings and a reranker from one API
Cohere
Jina AI - Generic embeddings miss the terms in your code, legal or finance documents
Voyage AI
- Text must not leave your servers, or embedding volume makes per-token pricing add up
Sentence Transformers
Compare the tools
Grouped by the side of the choice they answer, not ranked. Open a row for when to use it, the trade-off and what makers say.
Gemini APIapi · free tierApps on the Gemini API that want multilingual text embeddings on the same key.No evidence yet22.2M/wk4.9× vs npm
Use it whenGemini is already your main model provider.
Trade-offGeneral-purpose models, and like any hosted embedding, switching later means re-embedding everything.
In Lovable, ReplitNo maker's product we track shows it for this yet.
OpenAI APIapiApps already on the OpenAI API that want solid embeddings with no new vendor.No evidence yet37.4M/wk+60% vs npm
Use it whenYou already use OpenAI for text and your content is general-purpose English or common languages.
Trade-offGeneral-purpose models, with no reranking endpoint to pair them with.
llms.txtIn Lovable, Replit, RorkNo maker's product we track shows it for this yet.
CohereapiMultilingual search with embeddings and a reranker from the same vendor.


19+58 open source465k/wk−57% vs npm
Use it whenYour users search in many languages or you want reranking in one API.
Trade-offFree trial keys are rate-limited and not for commercial use.
Used by Chat Data, CustomGPT, Internet.io, Mnemosphere AI, Officely AI and 14 more · in 58 open-source projects.
We use Cohere as an endpoint in our team builder flow; you can combine this model with all available models.
Voyage AIapiRetrieval quality on specialized text, with models tuned for code, legal and finance documents.22 open source244.5k/wk+111% vs npm
Use it whenGeneric embeddings miss domain terms in your data.
Trade-offHosted API apart from one small open-weight model; now run by MongoDB, so its roadmap follows MongoDB's.
No maker's product we track shows it for this yet · in 22 open-source projects.
Jina AIapiMultilingual and long-document embeddings you can try without an API key.No evidence yet
Use it whenYou also need its Reader or reranker and want one vendor for the retrieval pipeline.
Trade-offKeyless use is rate-limited; now part of Elastic, so the product direction may shift.
No maker's product we track shows it for this yet.
Sentence Transformerslibrary · open sourceRunning open embedding and reranker models on your own CPU or GPU, with thousands of models on the Hugging Face Hub to choose from.125 open source5.3M/wk
Use it whenDocuments are private, volume is high, or you want to fine-tune an embedding model on your own data.
Trade-offPython only; you host, batch and scale the model yourself, and CPU-only servers embed slowly.
No maker's product we track shows it for this yet · in 125 open-source projects.
Who switches to what
Public pull requests on GitHub since Oct 2024 whose title says "X to Y" — real code changes, by developers in general rather than makers only. Pick a flow to see its pull requests.
- feat(dh): switch listup web search from OpenAI to Geminigrowthhackerssnu/admin · 2026-10-01
- Switch production thumbnails from OpenAI to Geminikevin4relocate/TripCaution · 2026-09-30
- feat: move OpenAI and Google defaults to GPT-6 and Gemini 3tag1consulting/ai-pr-review · 2026-09-29
- Migrate from OpenAI to Google Gemini for embeddings and LLManiilhr/agentic-ai-rag-chatbot · 2026-09-29
- fix: preserve image parts in OpenAI to Gemini responsesmoeacgx/NewAPI-Mao · 2026-09-27
- Updating from OpenAI to Gemini in llm.py following the potential toke…harpand87/banking-support-agent · 2026-09-27
- Switch the assistant LLM from OpenAI to Geministariik/Mia · 2026-09-26
- port(upstream): #4274 - disambiguate colliding tool call ids in openai to geminibloodf/durindoor · 2026-09-25
- feat(gemini): convert OpenAI Embeddings to Gemini embeddingstbphp/gpt-load · 2026-09-25
- Switch AI provider from OpenAI to Google GeminiOricrew/service-scheduler · 2026-09-24
- Switch Alora Chat from OpenAI to Google Gemini (free tier)vaniabhagania/Alora · 2026-09-24
- Podcast sources: switch from OpenAI to Google Gemininetteran/locreport · 2026-09-23
- fix(translator): disambiguate colliding tool call ids in openai to gemini requestdecolua/9router · 2026-09-23
- Switch the data analysis agent from OpenAI to Geminiluphone04/TalkToData_Data-Analytics-AI-Agent · 2026-09-21
- fix(relaykit): keep generated images when converting OpenAI chat to Geminimrdjango/models-gateway · 2026-09-20
- Switch analysis backend from OpenAI to Gemini APIgelson-ai/datachat-ai · 2026-09-18
- feat: bridge OpenAI Realtime to Gemini Livephongsathornpt/kokekokkor · 2026-09-16
- fix: changed chatbot ai api from openai to gemini due to insufficient…yani-clstn/Koolest · 2026-09-16
- Swap AI report generation from OpenAI to Google Geminirajibr0610/GRC-Quick-scan · 2026-09-14
- feat: translate OpenAI Responses to Geminiphongsathornpt/kokekokkor · 2026-09-13
- gemini fallback to openai fixesmaximhq/bifrost · 2026-10-02
- Switch LLM backend from Gemini to OpenAI gpt-4o-miniPiSquareLabs/PathShala · 2026-09-30
- perf(gemini): Migrate Gemini to OpenAI formattisfeng/Easydict · 2026-09-29
- fixes foreign-function-call-id issue for fallback from gemini to openaimaximhq/bifrost · 2026-09-29
- Switch the model calls from Gemini to OpenAI gpt-5.4-miniMishaMatcharashvili/trip-io · 2026-09-25
- refactor: migrate AI insights service from Gemini API to OpenAI APITkYaLa34/AI-Investment-Research-Platform- · 2026-09-23
- Switch AI provider from Gemini to OpenAI, add dialogue lessons, fix e…aaryashah1010/Fluentify · 2026-09-22
- migrated from gemini to openai llmmohitgarg196/CodeVitals-AI · 2026-09-21
- Fix Gemini adapter to support OpenAI SDK client formatarvindg4u/APIKeyRotator · 2026-09-20
- Fix Gemini chat model + image fallback to OpenAIfabelanorma3010/How--to-wite--a-fiction-book · 2026-09-05
- feat: migrate backend LLM from Gemini to OpenAI (OCR + chat)code-1705/ResolveAI · 2026-08-30
- Migrate BurilLab AI APIs from Gemini to OpenAIhaengjoo123/buril-lab · 2026-08-28
- fix(ts): a bare claude- or gemini- model no longer routes to OpenAIMervinPraison/PraisonAI · 2026-08-27
- Update LLM Advisor model documentation from Gemini to OpenAI gpt-5.4-nanodmboynton56/personal-portfolio · 2026-08-26
- Migrate Gemini providers to OpenAI, revamp dashboard UIsahasprajapati/daily-brief-editorial-ai · 2026-08-26
- fix(translator): preserve functionCall id in Gemini to OpenAI request translationdiegosouzapw/OmniRoute · 2026-08-24
- Switch link description generation from Gemini to OpenAIhectorcflores/my-reading-list · 2026-08-20
- Migrate AI provider from Gemini to OpenAI and update docstharakanst/cultural_food_guide · 2026-08-17
- Swap the production image provider from Gemini to OpenAIKyleMit/Splotch · 2026-08-14
- Migrate Day 11 runtime from Gemini to OpenAIQuocKhanhLuong/K3-Day-11-Guardrails-HITL-Responsible-AI · 2026-08-07
Before you choose
Pick a model by testing it on 50–100 real queries from your own data, not by public leaderboards. Record the model name and dimensions with every stored vector — switching models later means re-embedding everything. Adding a reranker on top of the first search often improves results more than changing the embedding model.
- Mixing vectors from two embedding models, or two dimensions of one model, in the same index; similarity scores between them are meaningless.
- Embedding whole documents as one vector, so a search can only ever return "the whole PDF" instead of the passage that answers the question.
- Choosing a model from a public leaderboard without testing it on your own queries and documents.
Decided alongside
What the 14 makers' products here chose for their other decisions.