Cohere embed-v4
by Cohere · paid · Last verified 2026-04-24
Cohere embed-v4 is a state-of-the-art multimodal embedding model supporting text and images in 100+ languages. It delivers top MTEB benchmark scores for multilingual retrieval and is optimized for enterprise RAG pipelines with support for int8 and binary quantization for efficient storage and search.
https://cohere.com/embeddings ↗C
C—Below Average
Adoption: C+Quality: B+Freshness: ACitations: CEngagement: F
Specifications
- License
- Proprietary
- Pricing
- paid
- Capabilities
- Integrations
- Use Cases
- API Available
- No
- Parameters
- Unknown
- Context Window
- 128K
- Modalities
- text, image
- Training Cutoff
- 2025
- Tags
- embeddings, multimodal, multilingual, rag, mteb, cohere
- Added
- 2026-04-24
- Completeness
- 60%
Index Score
44Adoption
50
Quality
70
Freshness
80
Citations
40
Engagement
0