[whatmodel.sh](../../README.md) / Ministral 8B (2410)
# Ministral 8B (2410)
Mistral · released 2024-10 · 8.02B params · mistral-research
**score 45.2** ±0.6 · #473 of 732 · from 39 benchmark results across 12 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Ministral 8B (2410) among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃●█▄▅▅▄▄▂▁ #265 of 313* 1.9% 3
coding █··········● #1 of 2* 2.9% 2
chat & writing ▁▁▁▃▆●▆█▆▇▇▃ #292 of 402 8.5% 4
math ▁▁▃●█▆▄█▄▂▁▁ #360 of 431 29% 5
agent tasks · #265 of 313 · 1.9% 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
Galileo Agent Leaderboard v1 (too… 68.9% ▃●▃▃▆▅▆▅█▂▅▂ beats 3/31 easy · avg 78% 0.9% at its level
BFCL v3 51.5% ▁▁▁▁▁▃●█▅▄▁▁ beats 41/114 medium · avg 51% 0.9% blunt · at its level
BFCL V4 (December 2025 archive) 11.1% ▁▁▁▂▆▆▅▆█▂▄▃ beats 0/114 hard · avg 31% — failed run: Same pattern as Nemotron Ul…
coding · #1 of 2 · 2.9% of its score
test result vs the others in its run difficulty counts why
HumanEval (maker) 76.8% ▁▁▁▁▂▂▆▆█●▂▁ beats 163/205 easy · avg 80% 1.5% sharp · at its level · noisy · small run
MBPP (maker) 70.0% ▁▁▁▄▅█▆●▄▁▂▁ beats 68/86 easy · avg 74% 1.4% at its level · small run
chat & writing · #292 of 402 · 8.5% of its score
Beats [Ministral 3 14B Reasoning 2512](ministral-3-14b-reasoning.md), [Ministral 3 8B](ministral-3-8b.md), [Gemma 4 E4B](gemma-4-e4b.md) here, though they rank higher overall. Trails [EXAONE 3.5 2.4B Instruct](exaone-3-5-2-4b.md), [LFM2 700M](lfm2-700m.md), [Flan-T5 XXL](flan-t5-xxl.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
LMArena (text) 1237.6 ▁▁▁▂▃●▃▆▆▇█▃ beats 87/385 avg 1260 4.1% blunt
AlpacaEval 2 length-controlled 31.4% ▃▃·▁▁●▇▅·▁·▁ beats 8/22 medium · avg 39% 2.2% sharp · at its level
Arena-Hard (version/judge unspeci… 70.9% ▃▃▂▅▃▆▅●▅▅▂▃ beats 21/30 easy · avg 70% 1.2% at its level · small run
IFEval prompt-level loose 56.4% ▁▃●▃▁▄▁▁█▃▃▁ beats 3/22 easy · avg 78% 0.9% at its level
math · #360 of 431 · 29% of its score
Beats [GPT-3.5 Turbo (0125)](gpt-3-5-turbo-0125.md), [DBRX Instruct](dbrx-instruct.md), [Command R Plus (04-2024)](command-r-plus-04-2024.md) here, though they rank higher overall. Trails [OPT-2.7B](opt-2-7b.md), [Gemma 2B (base)](gemma-2b-base.md), [Stable Code 3B](stable-code-3b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
GSM8K 80.0% ▂▃··▃▂●█▆▃▂▃ beats 7/22 very easy · avg 90% 11% sharp · several runs
MGSM 56.7% ▅▃▃●·▅···▃·▃ beats 5/10 easy · avg 72% 8.7% sharp · at its level
MATH level 5 14.9% ▂▂●▆▄█▆▃▅▂▃▄ beats 13/92 hard · avg 28% 4.0% sharp
MATH 41.6% ●▄██···▄▄▄·▄ beats 0/10 medium · avg 61% 3.0% at its level
MATH (subset unspecified) 40.0% ▃▃●▂▂▆·····▁ beats 7/22 medium · avg 58% 2.6% at its level · several runs
other tests (21)
test result vs the others in its run difficulty counts why
MMLU 68.5% ▁▂·▁·▁▁●·▃▃▅ beats 8/22 easy · avg 75% 13% at its level · several runs
BIG-Bench Hard 70.8% ▁▄·▁▁·▅●·▄▄▁ beats 13/22 medium · avg 72% 11% sharp · at its level · several runs
ARC Challenge 82.7% ▂▂▂●▂···▄··▂ beats 4/10 very easy · avg 89% 7.2% sharp
OpenBookQA 80.2% ▂··▄●▂·▂···▄ beats 5/10 easy · avg 85% 4.0% at its level
MMLU-Pro 36.6% ●▄▄·█▄▄▄·▄·▄ beats 1/10 medium · avg 54% 3.6% at its level
Multilingual MMLU (Microsoft) 53.7% █▄▄●█··█···▄ beats 4/10 medium · avg 63% 3.1% at its level
GPQA Diamond 27.1% ▁▁●▁▂▄▆▄▅▇█▅ beats 9/211 hard · avg 39% 2.5% sharp
BRIDGE clinical text, zero-shot a… 30.4% ▁▁▁▁▁▂▂▃●▆█▃ beats 32/97 hard · avg 31% 2.1% blunt · at its level
AraGen-v3 3C3H score (Inception, … 17.6% ▁▁·▁▂▂●█▄▅▄▄ beats 15/66 hard · avg 29% 1.6% typical
HellaSwag 73.5% ▃▅●·▃▃▃···▃▃ beats 3/10 easy · avg 82% 1.5% at its level
WinoGrande 63.1% ▃····▃●█▃·▅▅ beats 2/10 medium · avg 71% 1.4% at its level
Social IQA 77.6% ▃·▃▅▅█·●···▃ beats 9/10 medium · avg 76% 1.2% blunt · at its level
PopQA 20.2% ▁▁▁▃●▃▁█▁▃·▄ beats 6/21 hard · avg 31% 1.1% typical
BoolQ 80.5% ▃··▃··█●▃·▃▃ beats 6/10 easy · avg 83% 1.1% at its level
Open PT-LLM Leaderboard average (… 71.3% ▃▁▁▁▁▁▁▁▂●▅▁ beats 181/276 medium · avg 72% 1.0% at its level · noisy
GPQA (subset unspecified) 26.3% █··●▂··▂▂··▄ beats 5/10 very hard · avg 32% 0.9% sharp · too hard for it · noisy
EuroEval Life in the UK (English … 82.5% ▁▁▁▁▁▁▁▂▄●▄▃ beats 79/146 easy · avg 82% 0.6% at its level · noisy
PIQA 81.2% ▃·▅·▃█·●▅··▃ beats 7/10 medium · avg 82% 0.4% blunt · at its level
TruthfulQA MC2 55.5% ▃▃▂·▃●·▅█▂▅▂ beats 7/21 medium · avg 64% 0.3% blunt · at its level · several runs
Vectara hallucination leaderboard… 92.5% ▁▁▂▃●▄█▆▃▂▂▁ beats 39/140 very easy · avg 94% 0.3% blunt · maxed out · noisy
Vectara hallucination leaderboard… 92.6% ▁▂▁▂▅█▂●▄▃▁▁ beats 65/98 very easy · avg 90% 0.0% blunt
source results share of the score
microsoft/Phi-4-mini-instruct model card 16 55% ▇▇▇▇▇▇▇
2 OLMo 2 Furious 8 22% ▇▇▇
Epoch AI 2 6% ▇
Mistral AI vendor-reported benchmark sc… 4 5% ▇
LMArena leaderboard dataset 1 4% ▇
BRIDGE Medical Leaderboard 1 2% ▇
Inception Arabic/Hindi leaderboards 1 2% ▇
Open Portuguese LLM Leaderboard 1 1% ▇
Galileo Agent Leaderboard 1 1% ▇
Berkeley Function Calling Leaderboard r… 1 1% ▇
EuroEval leaderboards 1 1% ▇
Vectara hallucination leaderboard 2 0% ▇
prior: pull toward the middle 0% ▇
weighted average = 45.2
Licences: microsoft/Phi-4-mini-instruct model card ([MIT](../../licenses/card-microsoft-phi-4-mini-instruct.txt)) · 2 OLMo 2 Furious ([CC BY 4.0](../../licenses/paper-2501.00656v3-CC-BY-4.0.txt)) · Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · BRIDGE Medical Leaderboard ([MIT](../../licenses/bridge-medical-leaderboard-MIT.txt)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Galileo Agent Leaderboard ([Apache-2.0](../../licenses/galileo-agent-leaderboard-Apache-2.0.txt)) · Berkeley Function Calling Leaderboard result archive ([Apache-2.0](../../licenses/bfcl-result-archive-v4-Apache-2.0.txt)) · EuroEval leaderboards ([MIT](https://github.com/EuroEval/leaderboards/blob/1a77f1d2f8ba1ba05ab73287fb090d76bf41b2e0/LICENSE)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
Not counted (1)
source test score why
Berkeley Function Cal… BFCL V4 (December 2025 archiv… 11.1% failed run: Same pattern as Nemotron Ultra: zero on every category that needs a call and 100% on irrelevance, so the FC handler never got a parseable call.