Mistral AI
Best on the Capabilities Index Mistral Large 4 154.6
How does Mistral AI stack up?
Every model on Epoch AI's Capabilities Index by release date, with Mistral AI's in color.
Source: Epoch Capabilities Index, Epoch AI (CC BY 4.0)
Mistral AI's best models
Mistral Large 4
Flagship
154.6
#48 of 256
Epoch Capabilities Index
Self-Reported Index 158.4 #41 of 404
- Terminal-Bench 4.0 27.9% #19 /40
- GPQA Diamond 90.2% #48 /229
- GDP.pdf 32.8% #18 /19
Mistral Large 4 Preview
Open weights
152.7
#57 of 256
Epoch Capabilities Index
- GPQA Diamond 88.1% #64 /229
- Harvey LAB (Legal Agent Benchmark) 15.8% #10 /80
- Vals Finance Agent v2 54.7% #26 /80
Mistral Medium 3.5
Open weights Mid-size
141.6
#114 of 256
Epoch Capabilities Index
Self-Reported Index 143.5 #149 of 404
- IFBench 69% #46 /80
- Harvey LAB (Legal Agent Benchmark) 0.42% #61 /80
- Vals Finance Agent v2 32.1% #70 /80
Mistral Medium 3
Closed weights Mid-size
132.8
#151 of 256
Epoch Capabilities Index
Self-Reported Index 129.0 #283 of 404
- Humanity's Last Exam 4.52% #43 /47
- GPQA Diamond 59.5% #138 /229
- HiPhO 40.4% #24 /36
Mistral Large 2
Jul 2024 snapshot Open weights Flagship
126.3
#169 of 256
Epoch Capabilities Index
Self-Reported Index 126.9 #303 of 404
- GPQA Diamond 49% #159 /229
- KOR-Bench 53.1% #8 /33
- IHEval 29.4% #6 /10
Mistral Large
Closed weights Flagship
119.8
#185 of 256
Epoch Capabilities Index
Self-Reported Index 115.8 #362 of 404
- GPQA Diamond 38.8% #188 /229
- BigCodeBench-Hard 8.40% #86 /96
- CulturalBench 43.9% #9 /15
What benchmarks does Mistral AI mention most?
Here are the 28 benchmarks Mistral AI mentions most in its launch posts.
Source: Labs’ launch posts, read by Benchmark Almanac
MMLU Natural Questions TriviaQA MATH GSM8K HellaSwag WinoGrande PIQA ARC-Easy ARC-Challenge HumanEval MBPP BIG-Bench Hard AGIEval MT-Bench LMArena (Elo) AIME 2024 AIME 2025 GPQA Diamond LiveCodeBench v6 MATH-500 LiveCodeBench v5 Humanity's Last Exam MMMU MathVista MMMU-Pro SWE-bench Verified FLTEval
Source: Labs’ system cards and technical reports, read by Benchmark Almanac
Source: Labs’ launch posts, system cards and technical reports, read by Benchmark Almanac
Model cards
Every Mistral AI model
-
Mistral Large 4 (Le Chonk, open weights) Oct 2026
-
Voxtral Mini TTS 3 Oct 2026
-
Mistral Large 4 Oct 2026
-
Mistral Large 4 Preview Oct 2026
-
Leanstral 1.5 Jul 2026
-
Mistral Medium 3.5 Apr 2026
-
Leanstral Mar 2026
-
Mistral Small 4 Mar 2026
-
Devstral 2 Dec 2025
-
Devstral Small 2 Dec 2025
-
Ministral 3 14B Dec 2025
-
Ministral 3 3B Instruct Dec 2025
-
Ministral 3 8B Instruct Dec 2025
-
Mistral Large 3 Dec 2025
-
Magistral Medium 1.2 Sep 2025
-
Magistral Small 1.2 Sep 2025
-
Mistral Medium 3.1 Aug 2025
-
Magistral Small 1.1 Jul 2025
-
Voxtral Small Jul 2025
-
Devstral Medium Jul 2025
-
Devstral Small 1.1 Jul 2025
-
Mistral Small 3.2 Jun 2025
-
Magistral Medium 1.0 Jun 2025
-
Magistral Medium 1.1 Jun 2025
-
Magistral Small 1.0 Jun 2025
-
Devstral Small 1.0 May 2025
-
Mistral Medium 3 May 2025
-
Mistral Small 3.1 Mar 2025
-
Mistral Saba Feb 2025
-
Mistral Small 3 Jan 2025
-
Codestral Jan 2025
-
Pixtral Large Nov 2024
-
Mistral Large 2 Nov 2024 Nov 2024
-
Ministral 3B Oct 2024
-
Ministral 8B Oct 2024
-
Mistral Small v24.09 Sep 2024
-
Pixtral 12B Sep 2024
-
Mistral Large 2 Jul 2024 Jul 2024
-
Mistral NeMo Jul 2024
-
Mistral-Nemo-Base-2407 Jul 2024
-
Mathstral-7B-v0.1 Jul 2024
-
Mixtral 8x22B Apr 2024
-
Mistral Large Feb 2024
-
Mistral-7B-Instruct-v0.2 Dec 2023
-
Mistral 7B v0.2 Dec 2023
-
Mistral Medium Dec 2023
-
Mixtral 8x7B Dec 2023
-
Mistral-7B-Instruct-v0.1 Sep 2023
-
Mistral 7B v0.1 Sep 2023
-
Mistral 7B v0.3 Sep 2023