whatmodel.sh / Command A Vision (07-2025)

Command A Vision (07-2025)

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Command A Vision (07-2025) is an open-weights language model from Cohere.

Catalogue identity notes

HF repo sha e1016a8105950a626e7dd91b0a3030f7ae522411, cc-by-nc-4.0, created 2025-07-28; 112B per the Cohere post's comparison table.

56.7±3.8comparison only: no current run option verifiedfrom 2 benchmark results across 1 source* thinly tested, treat the rank as provisional

$ how it compares

The LLMs ranked either side of Command A Vision (07-2025), out of 1132 with a score, then more from Cohere.

#modelscorewhereRAMreleased
365Llama 4 Maverick (402B-A17B) Meta57±1api $0.42/M +local~65GB2025-04
366Claude 3.5 Sonnet (2024-06-20) Anthropic57±1comparison only2024-06
367Command A Vision (07-2025) Cohere57±4comparison only
368Qwen2.5 Coder 32B Instruct Alibaba57±1api $0.74/M +local~21GB2024-11
369Qwen3 4B Thinking 2507 Alibaba57±1local~3.9GB2025-08
more from Cohere
85North Small Translate 1.0 Cohere71±6api price unknown2026-08
205Command A+ (05-2026) Cohere64±1api price unknown +local~71GB2026-05
357Command A (03-2025) Cohere57±1api $4.38/M +local~37GB2025-03

see all LLMs · find the ones that fit your machine

$ what it's good at

vision & documents█··········█ #2 of 2 * · 94% of its score · 2 tests
testresultvs the others in its rundifficultycountswhy
CharXiv descriptive questions lab82.0% ✎▂▄▄▅▇▄█▇▇▂▁▂ beats 27/58easy · avg 71%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 51%small run
CharXiv (reasoning questions) lab46.9% ✎▂▅▃▅▅▆▆▅▆█▆▂ beats 25/100hard · avg 34%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 43%at its level · small run

$ where to run it

No verified API or download recorded yet · add one

Weights and documentation: huggingface.co/CohereLabs ↗

$ how the score was worked out · Cohere vendor-reported benchm… 94%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
Cohere vendor-reported benchmark scores for its own models (cohere.com blog) · lab-reported2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 94%Factual vendor-reported score (not licensed content)
prior · pull toward the middle▇▇ 6%
weighted average= 56.7

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json