Cohere Embed v4

Cohere

Cohere's multimodal embedding model, a top MTEB performer that embeds text and images into a shared space with support for 100+ languages and long documents. Built for enterprise search and RAG, it pairs with Cohere Rerank for high-precision retrieval pipelines.

Strengths

Top MTEB performance (~65.2) among leading models
Multimodal: embeds text and images in a shared space
100+ languages with consistent quality
Long-document support (up to ~128K tokens)
Enterprise search/RAG, pairs with Cohere Rerank

Caveats

Proprietary - API and private deployment (no open weights)
Usage-based pricing scales with volume
Best value realized with the Cohere retrieval stack

Vendor: Cohere
Context Window: 128,000 tokens
License: Proprietary
Modalities: text
image

Capabilities

Vision

Audio

Video

Tool Use

Resources

API Documentation

Pricing

Usage-based via the Cohere API (per 1M tokens / per image); available on major clouds and for private deployment.