Embedding Model Lab
Compare embedding models on your own text. Because switching models means re-embedding your whole corpus, the time to compare is before you commit. Bring your own provider keys — they stay in your browser.
Provider keys
Stored only in your browser (localStorage). Sent only with your embedding requests, never saved on our servers.
Your data
Models (2 selected)
Add at least 2 documents to compare.
No key for OpenAI, Voyage — those models will be skipped.
Compare embedding models side-by-side
Add a provider key, paste some documents, pick a few models, and run. Or start with a built-in sample corpus.