HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/mistralai/mistralai/Ministral-3-8B-Reasoning-2512

mistralai/Ministral-3-8B-Reasoning-2512

family mistral · 7 variants · unknown

provider
mistralai
released
—
huggingface
https://huggingface.co/mistralai/Ministral-3-8B-Reasoning-2512
best scores
AIME 2025: 73.7AIME 2024: 83.7GPQA Diamond: 66.3LiveCodeBench: 59.3Chatbot Arena: 42.7WildBench: 65.1MATH: 87.0MT-Bench: 6.7MMLU: 74.2AGIEval: 64.8TriviaQA: 74.9

Best scores per benchmark

benchmarkcategorybest score
MATHmath87.0
AIME 2024math83.7
TriviaQAknowledge74.9
MMLUknowledge74.2
AIME 2025math73.7
GPQA Diamondreasoning66.3
WildBenchchat65.1
AGIEvalreasoning64.8
LiveCodeBenchcoding59.3
Chatbot Arenachat42.7
MT-Benchchat6.7

Variants (7)

variantkindquantbitsvRAMdlorgyear
Ministral-3-8B-Reasoning-2512-GGUFggufgguf16.05,954unsloth2025
Ministral-3-8B-Reasoning-2512-GGUFggufgguf4.03,487lmstudio-community2025
Ministral-3-8B-Reasoning-2512-GGUFggufgguf16.02,369mistralai2026
mistralai_Ministral-3-8B-Reasoning-2512-GGUFggufgguf2.01,185bartowski2025
Ministral-3-8B-Reasoning-2512-unsloth-bnb-4bitbnb-4bit4bit266unsloth2025
Ministral-3-8B-Reasoning-2512-GGUFggufgguf8.0235ggml-org2025
Ministral-3-8B-Reasoning-2512originalbf1616.069unsloth2025
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.