{"id":"cohere-rerank-5","slug":"cohere-rerank-5","name":"Cohere Rerank 5","orgId":"cohere","family":"cohere-rerank-5","releaseDate":null,"availability":"announced","openWeights":false,"license":null,"paramsTotal":null,"paramsSource":null,"contextWindow":null,"knowledgeCutoff":null,"modelCardUrl":"https://cohere.com/blog/rerank-5","hfRepo":null,"modalities":["text"],"aliases":["Cohere Rerank 5","rerank-v5.0-pro","rerank-v5.0-fast","cohere rerank 5","cohere rerank 5 pro","cohere rerank 5 fast","rerank 5 pro","rerank 5 fast"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-10-21","announced":"2026-10-21","blurb":"Cohere's fifth-generation reranking models for search, RAG and agents, in Pro (highest quality) and Fast (low latency, high throughput) tiers. New in this version: Context Focus, which returns the most relevant passages inside each document in the same call, and calibrated relevance scores. Text input, 32K context, 100+ languages. API only, no open weights.","api":"live","apiModelId":"rerank-v5.0-pro","openWeights":"no","params":{"text":"About 2B (Pro), about 500M (Fast)","source":"Cohere Rerank documentation"},"facts":[{"label":"Models","value":"rerank-v5.0-pro and rerank-v5.0-fast"},{"label":"Context","value":"32,768 tokens, query and document combined"},{"label":"Output","value":"Documents ranked by relevance with a calibrated 0–1 score; optional Context Focus returns scored character spans"},{"label":"Input","value":"Text; JSON, tables and records passed as YAML strings; 100+ languages"},{"label":"Price","value":"$0.05 per 1M tokens (Pro), $0.02 per 1M tokens (Fast)"},{"label":"Availability","value":"Cohere API, Model Vault, Microsoft Foundry, Amazon SageMaker, North, and private VPC or on-premises deployments"}],"claims":["Lab claim: Rerank 5 Pro scores 82.2 on BRIGHT, 5.8 points above Voyage AI Rerank 3 and 12.4 above Rerank 4 Pro, and was first on 39 of 46 composite datasets in Cohere's evaluation (RCP-nDCG@10, a Cohere-developed metric).","Lab claim: Rerank 5 Fast reranks about 187 documents of 1,000 tokens per second on one NVIDIA A100, 2.9× Pro's throughput, and beats Rerank 4 Pro."],"sources":[{"kind":"lab","title":"Introducing Rerank 5 (Cohere blog)","url":"https://cohere.com/blog/rerank-5","date":"2026-10-21"},{"kind":"lab","title":"Rerank documentation (Cohere)","url":"https://docs.cohere.com/docs/rerank","date":"2026-10-21"},{"kind":"lab","title":"Cohere announcement on X","url":"https://x.com/cohere/status/2112923503100125326","date":"2026-10-21"}],"firstBenchmark":null},"organization":{"id":"cohere","name":"Cohere","slug":"cohere","country":null,"url":null,"color":"#39594D","mark":"C"},"price":null,"answer":{"word":"Pending benchmarks.","scope":"Announced Oct 21, 2026; no independent results yet.","lead":"As of Oct 22, 2026, Cohere Rerank 5 is pending benchmarks on isitsota: announced Oct 21, 2026, with no independent results on the boards we track yet.","confidence":"low"},"status":"PENDING_BENCHMARKS","verdicts":[]}