Retrieval & RAGbeginner11 min · 10 steps
Embedding Providers
OpenAI, Gemini, Cohere, Jina, Nomic, Ollama and RAG platforms all turn text into vectors — they differ in hosting, languages, modalities and how they treat queries versus documents.
dataerrorrequestresponse
Key ideaStep 1 of 10
Every provider builds its own space
An embedding model maps text to a point in a space it learned. Two providers produce vectors of different sizes and meanings — a query embedded by one is meaningless against an index built by another.
So the provider choice is a commitment: the same model must embed every document and every query, and switching means re-embedding the corpus.
Watch outStore the model name and version next to every vector. Mixed indexes fail silently — they still return neighbours, just the wrong ones.
What changed
- +Two providers, two incompatible spaces
Lesson 21 of 37 in Retrieval & RAG