Source: Cohere embeddings documentation
Cohere's guide to its Embed models (embed-v5.0-pro and embed-v3.0): input_type, Matryoshka dimensions, compressed embedding types and image embeddings.
Contents
- What: API documentation
- Author: Cohere
- Date: living document (read 2026-11-14)
- Original: docs.cohere.com/docs/embeddings
Summary
As of November 2026 the page covers embed-v5.0-pro, a multilingual multimodal model, and the older embed-v3.0 family. Cohere makes the task explicit through input_type, offers Matryoshka dimensions, and can return several compressed representations in one call.
Key claims
input_type:search_query,search_document,classification,clustering(andimagefor v3.0); it tailors the embedding to its role.embed-v5.0-prosupports output dimensions 256, 512, 768, 1024, 1536 and 2048.embedding_types:float,int8,uint8,binary,ubinary; several can be requested at once. A 1024-dim binary vector is 128 bytes and may need unpacking for some vector stores.embed-v5.0-procan embed text and images together in one vector (e.g. slides, document screenshots); up to 96 images per request.