Benchmark Almanac

Cohere

16 models 10 model cards read cohere.com

What benchmarks does Cohere mention most?

Here are the 15 benchmarks Cohere mentions most in its launch posts.

Launch posts System cards All reports

MMLU IFEval tau-bench Berkeley Function Calling Leaderboard (BFCL) MMMU ChartQA InfographicVQA AI2D MathVista DocVQA OCRBench RealWorldQA WMT24++ Terminal-Bench Hard SciCode

Source: Labs’ launch posts, read by Benchmark Almanac

Model cards

Date	Model	Document	Benchmarks
Sep 10, 2026	North Small Translate 1.0	Introducing North Small Translate: A leading sovereign open-weight machine translation model (Cohere blog)	1
Aug 12, 2026	North Micro Vision Instruct	North Micro Vision Instruct model card (Hugging Face)	20
Jun 9, 2026	North Mini Code 1.0	Introducing North Mini Code: Cohere's first model for developers (Cohere blog)	6
May 20, 2026	Command A+	Introducing Command A+: Making sovereign agentic capabilities available to all (Cohere blog)	12
Mar 26, 2026	Cohere Transcribe	Introducing Cohere Transcribe: a new state-of-the-art in open-source speech recognition (Cohere blog)	2
Aug 28, 2025	Command A Translate	Introducing Command A Translate (Cohere blog)	1
Aug 21, 2025	Command A Reasoning	Command A Reasoning: Enterprise-grade control for AI agents (Cohere blog)	4
Jul 31, 2025	Command A Vision	Command A Vision: Multimodal AI for Business (Cohere blog)	9
Mar 13, 2025	Cohere Command A	Introducing Command A: Max performance, minimal compute (Cohere blog)	8
Mar 4, 2025	Aya Vision 32B	Cohere For AI Launches Aya Vision, a Multilingual Multimodal AI Model (Cohere blog)	2

Every Cohere model