[whatmodel.sh](../../README.md) / Command R7B (12-2024)
# Command R7B (12-2024)
Cohere · released 2024-12 · 8.03B params · cc-by-nc-4.0
**score 45.5** ±1.3 · #463 of 732 · from 14 benchmark results across 4 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Command R7B (12-2024) among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃●█▄▅▅▄▄▂▁ #261 of 313 19% 3
coding ▁▁▁▃●▇█▇▅▄▃▁ #375 of 469 35% 5
chat & writing ▁▁▁▃▆▄▆●▆▇▇▃ #231 of 402* 12% 2
agent tasks · #261 of 313 · 19% of its score
Beats [DeepSeek R1 0528 Qwen3 8B](deepseek-r1-0528-qwen3-8b.md), [Magistral Small 1.0 (2506)](magistral-small-2506.md), [Llama 4 Scout (109B-A17B)](llama-4-scout.md) here, though they rank higher overall. Trails [Granite 4.0 H 1B](granite-4-0-h-1b.md), [Granite 4.0 1B](granite-4-0-1b.md), [SmolLM3 3B](smollm3-3b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Verified (maker) 3.6% ●···▃·▃·▃█·▃ beats 0/7 hard · avg 17% 9.3% sharp
BFCL V4 (December 2025 archive) 32.1% ▁▁▁▂█●▅▅▆▂▂▂ beats 35/77 hard · avg 31% 6.0% typical
BFCL v3 51.7% ▁▁▁▁▁▃●█▅▄▁▁ beats 42/114 medium · avg 51% 3.7% blunt · at its level
coding · #375 of 469 · 35% of its score
Beats [LFM2.5 8B-A1B](lfm2-5-8b-a1b.md), [Qwen3-VL 2B Thinking](qwen3-vl-2b-thinking.md), [DeepSeek R1 Distill Qwen 7B](deepseek-r1-distill-qwen-7b.md) here, though they rank higher overall. Trails [Qwen2.5 1.5B Instruct](qwen2-5-1-5b.md), [WizardLM 13B V1.2](wizardlm-13b-v1-2.md), [Code Tulu 2 13B](code-tulu-2-13b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
MBPP+ (maker) 72.0% ●█·██···████ beats 0/7 easy · avg 78% 10% at its level
SWE-bench Verified (maker) 3.6% ●···▃·▃·▃█·▃ beats 0/7 hard · avg 17% 9.3% sharp
LiveCodeBench v5 (Oct 2024–Feb 20… 9.0% ●▄·▄·····█▄█ beats 0/7 hard · avg 18% 6.6% typical
BIRD-SQL (dev) (maker) 42.2% ●··▄·▄·█··▄█ beats 0/7 medium · avg 47% 5.6% at its level
Aider polyglot (225 exercises, tw… 2.7% █●··█·█····▄ beats 2/7 very hard · avg 5.5% 3.3% sharp · too hard for it · noisy
chat & writing · #231 of 402 · 12% of its score
Beats [Granite 4.2 8B](granite-4-2-8b.md), [Mistral Small 4 (119B-A6.5B)](mistral-small-4.md), [Olmo 3.1 32B Think](olmo-3-1-32b-think.md) here, though they rank higher overall. Trails [Flan-T5 XXL](flan-t5-xxl.md), [LFM2.5 1.2B Instruct](lfm2-5-1-2b-instruct.md), [Gemma 3 4B](gemma-3-4b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Arabic IFEval prompt-level (Incep… 56.4% ▁·▁▁▂▁▄▄▆▄●▄ beats 30/38 medium · avg 46% 12% sharp · at its level
IFEval strict, mean of prompt and… 77.9% ●··········█ beats 0/1 — not counted yet: no other model in the …
other tests (7)
test result vs the others in its run difficulty counts why
RepoQA (32K, threshold 0.8) (make… 69.6% ●▄▄··▄▄··▄·█ beats 0/7 easy · avg 80% 15% at its level
Open Persian LLM Leaderboard v2 a… 50.4% ▂▄▂▃●▄▃▆▃▅▂▂ beats 30/81 medium · avg 57% 7.8% blunt · at its level
BigCodeBench (Instruct) (maker) 30.9% ▄··●▄···▄▄█▄ beats 1/7 hard · avg 35% 6.4% at its level
MMLU-Pro (maker) 42.4% ▁▁·▁▁▁●▂▄▅█▁ beats 33/188 medium · avg 54% 6.1% at its level · small run
GPQA Diamond (simple-evals) (make… 26.3% ●·····▄█▂▂▃▃ beats 0/22 very hard · avg 35% 5.2% sharp · too hard for it · small run
MMLU (simple-evals, zero-shot CoT… 65.2% ●···▂▁▅█▂▂▃▁ beats 0/22 easy · avg 76% 3.0% at its level · small run
InFoBench (maker) 85.6% ●··········█ beats 0/1 — not counted yet: no other model in the …
source results share of the score
Cohere (submitted / maker-reported) 10 70% ▇▇▇▇▇▇▇▇
Inception Arabic/Hindi leaderboards 1 12% ▇
Berkeley Function Calling Leaderboard r… 2 10% ▇
Open Persian LLM Leaderboard v2 1 8% ▇
prior: pull toward the middle 1% ▇
weighted average = 45.5
Licences: Cohere ([CC BY 4.0](../../licenses/report-2504.00698v2.txt)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Berkeley Function Calling Leaderboard result archive ([Apache-2.0](../../licenses/bfcl-result-archive-v4-Apache-2.0.txt)) · Open Persian LLM Leaderboard v2 ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt))
Not counted (2)
source test score why
Cohere IFEval strict, mean of prompt… 77.9% too few models have it yet
Cohere InFoBench 85.6% too few models have it yet