Rerank 5 Fast
Rerank 5 Fast offers the lighter tier of Cohere's current reranking family. It retains Context Focus and calibrated relevance scoring for applications that prioritize throughput and response time.
Model behavior
Rerank operates on a supplied candidate list. It does not require that list to come from a particular embedding model. The current models are available through Cohere, private deployment options and supported platforms; OpenRouter access is also documented.Embed 5 and Rerank 5 on OpenRouter
How to use it well
Practical guidance. Fast fits a search box or an agent that makes many searches during a task. Compare the final visible results with and without reranking, and include response time in that review. If the correct source is absent from the candidate list, improve retrieval before tuning the reranker. Context Focus is most useful when a long source contains a small amount of evidence relevant to the current question.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | rerank-v5.0-fast |
| Context window | 32,768 tokens Context is shared by the query and each candidate document. Cohere Rerank specifications |
| Model scale | Approximately 500 million parametersCohere Rerank specifications |
| Input modality | TextCohere Rerank specifications |
Related models
- Rerank 5 Pro
Cohere · Rerank
Quality-focused search reranker with Context Focus for returning relevant document passages.
Launch date not confirmedAPI - Embed 5 Fast
Cohere · Embed
Low-latency embedding model that shares its vector space with Embed 5 Pro.
Launch date not confirmedAPI - Rerank 4 Fast
Cohere · Rerank
Throughput-focused multilingual document reranker in the Rerank 4 family.
Launch date not confirmedAPI