Cohere Rerank 5 Fast

Not yet independently verified. Rests on Cohere's own announcement of 21 October 2026. The benchmark results are Cohere's own, measured with a metric Cohere developed, and no independent dated test was found; the page states no context length or licence. We will update this when it can be confirmed, and remove this note.

Cohere Rerank 5 Fast is the speed-optimised reranking model in the Rerank 5 generation that Cohere released on 21 October 2026, next to Rerank 5 Pro. It reorders documents returned by a first search so that the most relevant reach a language model or a user first, at a lower price and latency than the Pro model.

What it does

Like Pro, it returns calibrated relevance scores and supports Cohere's Context Focus option, which returns the most relevant passages within each document rather than whole documents. Cohere says that in all four of its headline benchmarks Rerank 5 Fast outperforms Voyage AI Rerank 3 Lite, Qwen3-Reranker-0.6B, the larger Qwen3-Reranker-8B and Cohere's previous Rerank 4 Pro. Those results are Cohere's own, measured with its RCP-nDCG@10 metric.

Who it is for

Rerank 5 Fast is aimed at high-volume search, chat and agent pipelines where every query is reranked and the cost per call adds up. At $0.02 per million tokens it costs 40 percent of the Pro price. It is available on the same platforms as Pro: the Cohere API and Model Vault, Microsoft Foundry, Amazon SageMaker and Cohere's North, and for private VPC or on-premises deployment.

Key facts

Maker: Cohere Released: 21 October 2026 Price: Pro $0.05, Fast $0.02 per 1M tokens on the Cohere API Available on: Cohere API, Model Vault, Microsoft Foundry, Amazon SageMaker, North Licence: not stated on the announcement page Variant: Fast, the speed-optimised tier