What is an embedding model?
What is it?
An embedding model turns a text into a list of numbers (a vector). Texts with similar meaning give close vectors: it is the basis of semantic search and RAG.
When to use it
- Pick the model of a RAG by language, document size and score.
- Find the right settings: prefixes, distance, chunk size.
Page fields
| Model | Hugging Face identifier or API name. |
|---|---|
| Dimensions | Size of the vector. |
| Max tokens | Maximum length of a text. |
| Distance | Cosine, dot product or Euclidean. |
| Prefixes | Text to put before queries and passages, if the model needs it. |
| Benchmarks | Published scores, e.g. MTEB. |
How to use it
The Python example loads the model with sentence-transformers and applies the right prefixes.