Mistral Large 3: pricing, context window and benchmarks
Mistral Large 3 is a multimodal model from Mistral AI. It offers a 128,000-token context window, costs $5.41 per million input tokens and $23.83 per million output tokens, and averages 81.7 across the index's benchmark suite.
Pricing
Mistral Large 3 costs $5.41 per million input tokens and $23.83 per million output tokens. A blended 3:1 input-to-output mix works out to about $10.02 per million tokens. Cached input is listed at $0.60 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $5.41 | per 1M tokens |
| Output | $23.83 | per 1M tokens |
| Cached input | $0.60 | per 1M tokens |
| Blended 3:1 | $10.02 | estimate per 1M tokens |
| 1M in + 1M out | $29.24 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 84.6 | 71 | 88th | |
| GPQA Diamond | 79.6 | 65 | 89th | |
| SWE-bench Verified | 71.3 | 89 | 85th | |
| HumanEval | 88.6 | 62 | 90th | |
| MATH-500 | 84.1 | 123 | 80th | |
| AIME 2025 | 78.3 | 110 | 82th | |
| LiveCodeBench | 74.4 | 101 | 83th | |
| IFEval | 89.8 | 75 | 88th | |
| TruthfulQA | 83.7 | 89 | 85th | |
| HellaSwag | 88.2 | 135 | 78th | |
| MMMU | 76.2 | 134 | 78th | |
| Arena Elo | 1448.6 | 76 | 87th |
What Mistral Large 3 is good for
Mistral Large 3 is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $5.41 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- No standout weakness in the indexed specifications, but real-world behaviour still needs testing.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| o3 Vision | OpenAI | 1000000 | $15.24 | 81.8 |
| Llama 4 Maverick Instruct | Meta AI | 200000 | $5.33 | 81.7 |
| Mistral Large 3 | Mistral AI | 128000 | $5.41 | 81.7 |
| Claude 3.7 Sonnet | Anthropic | 200000 | $14.91 | 81.6 |
| Gemini 2.5 Pro | Google DeepMind | 200000 | $15.20 | 81.6 |
| Llama 4 Maverick | Meta AI | 200000 | $5.36 | 81.6 |
FAQ
How much does Mistral Large 3 cost?
Mistral Large 3 is listed at $5.41 per million input tokens and $23.83 per million output tokens. Cached input is $0.60 per million tokens. A blended 3:1 input-to-output mix is about $10.02 per million tokens.
What is the context window of Mistral Large 3?
Mistral Large 3 has a context window of 128,000 tokens and can emit up to 4,096 output tokens per request in this index.
Is Mistral Large 3 open weights?
Yes. Mistral Large 3 is listed as an open-weight model under a open weights licence.
How does Mistral Large 3 score on benchmarks?
Mistral Large 3 averages 81.7 across the scored benchmarks in this index, ranking #90 of 600 models. Its strongest relative result is Arena Elo at 1448.6.
What is Mistral Large 3 best for?
Mistral Large 3 is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $5.41 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.