Context management
Cohere Rerank
Retrieval-optimization model API that reranks agent-retrieved documents before they reach the LLM
Commercial
What it is
Our note
Reranking is the cheapest known lever on agent context quality, and Cohere bills it per search unit, defined on its pricing page as one query plus up to 100 documents, with long-document chunks counted separately. Distinct from the Cohere North 2 agent-memory product already in the registry: this entry is the retrieval layer.
At a glance
- Vendor
- Cohere
- Business model
- Commercial
- Self-hosting
- No, hosted only
- Access
- REST /v2/rerank endpoint; rerank-v4.0-pro and rerank-v4.0-fast; private/VPC deployment options
Links and sources
- Official sitecohere.com/rerank
- Pricingcohere.com/pricing
- Sourcedocs.cohere.com/docs/rerank
- Sourcecohere.com/blog/rerank-5
Facts verified nov 10, 2026; added nov 10, 2026. Vendor facts come from the vendor’s own pages; benchmark claims are self-reported unless we say otherwise.