Mistral Large 3: benchmark scores, price and ranking
Mistral Large 3 from Mistral AI ranks #697 of 743 LLMs on the LLMs Tiger aggregate benchmark score (42.7), based on 6 benchmark results captured up to 2026-10-13.
- Organisation
- Mistral AI
- Type
- Frontier model
- Open weights
- No
- Released
- 2025-12-02 (source: epoch)
- Cheapest API price
- $0.500 in / $1.50 out per million tokens (via azure_ai, observed 2026-10-13T20:47Z)
- Aggregate rank
- #697 of 743 (score 42.7)
Relative to other models, Mistral Large 3 is strongest on DTBench (#83 of 146) and weakest on WebDev Arena (#62 of 67).
Across the 6 benchmarks where at least ten models are scored, Mistral Large 3 is in the top 10% on 0 and in the top half on 0.
Its cheapest listed input price, $0.500 per million tokens, is lower than that of 48% of the 427 models with a listed price.
Mistral Large 3 benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| DTBench | 65.1% | #83 of 146 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 36.2% | #75 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
| LMCA | 16.7% | #82 of 108 | Epoch AI Benchmarking Hub | 2026-10-13 |
| CritPt | 0.0% | #105 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ALE-Bench | 265 | #73 of 79 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WebDev Arena | 1,230 | #62 of 67 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Mistral Large 3
Mistral Large 3 is a frontier model from Mistral AI. It was released on 2025-12-02, according to Epoch AI. The cheapest API listing we track charges $0.500 per million input tokens and $1.50 per million output tokens, through azure_ai. We hold 6 benchmark results for it across 4 areas, and none is in the top tenth of its board.
Where Mistral Large 3 stands
On our aggregate score, Mistral Large 3 is #697 of 743 ranked models, in the lower half. Its score is 42.7; the scale is centred on 50, so it sits 7.3 points below the typical model. That beats 6% of the ranked field. Among the 44 ranked Mistral AI models it is #39. The next one up is mistral-large-2402 at 42.8. The next one down is ministral-8b-2410 at 42.6.
Mistral Large 3, benchmark by benchmark
Mistral Large 3 lands below the middle on DTBench: 65.1%, #83 of 146, with 56% of the field scoring higher. Claude Opus 5.5 (max) leads at 98.9%; the closest result above is grok-2-1212 at 65.2%. It covers dynamic tool-use benchmark. The board median is 68.8%, so Mistral Large 3 is 3.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Mistral Large 3 lands below the middle on SciCode: 36.2%, #75 of 106, with 70% of the field scoring higher. Claude Opus 5.5 (max) leads at 66.9%; the closest result above is Qwen3.6 27B at 37.3%. It covers scientific research code. The board median is 44.3%, so Mistral Large 3 is 8.1 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Mistral Large 3 lands below the middle on LMCA: 16.7%, #82 of 108, with 75% of the field scoring higher. Claude Opus 5.5 (max) leads at 68.2%; the closest result above is claude-3-opus-20240229 at 17.0%. It covers language-model capability assessment. The board median is 29.2%, so Mistral Large 3 is 12.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
CritPt is a weak spot. Mistral Large 3 scores 0.0%, #105 of 115, behind 90% of the models measured. The leader, GPT-5.6 Sol (max), reaches 32.3%. It covers unpublished research-level physics problems. The board median is 2.9%, so Mistral Large 3 is 2.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
ALE-Bench is a weak spot. Mistral Large 3 scores 265, #73 of 79, behind 91% of the models measured. The leader, GPT-6 Astra (max), reaches 2,951. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so Mistral Large 3 is 499 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
WebDev Arena is a weak spot. Mistral Large 3 scores 1,230, #62 of 67, behind 91% of the models measured. The leader, Claude Opus 5.5 (max), reaches 1,820. It covers building web apps, crowd-voted (Elo). The board median is 1,433, so Mistral Large 3 is 204 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Software engineering: weak, ahead of 15% of the field on average across 3 benchmarks. Its best board here is SciCode (#75 of 106) and its weakest is WebDev Arena (#62 of 67).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Mistral Large 3 costs about the same as the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on Mixtral-8x22B-Instruct-v0.1 ($0.600) and gemma-2-27b ($0.650).
Release history
The Mistral AI model released before it was Magistral Small 1.2 on 2025-09-18, which scores 43.8. The next one was Devstral Small 2 on 2025-12-09, at 45.4.
Reading these numbers
The 6 scores on this page come from 1 source: Epoch AI Benchmarking Hub (6). No official leaderboard supplied any of them, so treat them as third-party measurements. With 6 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #694 mistral-large-2402 (42.8)
- #695 gemini-1.5-flash-002 (42.8)
- #696 deepseek-llm-67b-chat (42.8)
- #698 granite-4.0-1b (42.7)
- #699 Qwen2.5-Coder-0.5B-Instruct (42.6)
- #700 Llama-3.1-70B-Instruct (42.6)
More from Mistral AI
- Mistral Large 2 (2407) (#231, 49.9)
- codestral-2405 (#352, 48.4)
- pixtral-large-2411 (#360, 48.3)
- Codestral-22B-v0.1-Q4_K_M (#392, 48.0)
- codestral:22b-v0.1-q8_0 (#393, 48.0)
- Mistral Large 4 (high) (#411, 47.7)
Compare it on the live leaderboard → · How the aggregate score works