Mistral Large 3 Instruct: pricing, context window and benchmarks
Mistral Large 3 Instruct is a multimodal model from Mistral AI. It offers a 400,000-token context window, costs $5.31 per million input tokens and $22.47 per million output tokens, and averages 81.1 across the index's benchmark suite.
Pricing
Mistral Large 3 Instruct costs $5.31 per million input tokens and $22.47 per million output tokens. A blended 3:1 input-to-output mix works out to about $9.60 per million tokens. Cached input is listed at $0.63 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $5.31 | per 1M tokens |
| Output | $22.47 | per 1M tokens |
| Cached input | $0.63 | per 1M tokens |
| Blended 3:1 | $9.60 | estimate per 1M tokens |
| 1M in + 1M out | $27.78 | 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 | 82.3 | 116 | 81th | |
| GPQA Diamond | 79.8 | 60 | 90th | |
| SWE-bench Verified | 71.9 | 76 | 87th | |
| HumanEval | 84.2 | 129 | 79th | |
| MATH-500 | 83.4 | 140 | 77th | |
| AIME 2025 | 76.2 | 137 | 77th | |
| LiveCodeBench | 75.1 | 94 | 84th | |
| IFEval | 87.6 | 113 | 81th | |
| TruthfulQA | 85.5 | 60 | 90th | |
| HellaSwag | 89.1 | 110 | 82th | |
| MMMU | 76.5 | 125 | 79th | |
| Arena Elo | 1446.2 | 88 | 85th |
What Mistral Large 3 Instruct is good for
Mistral Large 3 Instruct 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 $5.31 per million input tokens, it sits in the mid-range of the index, and the 400,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 400,000-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 |
|---|---|---|---|---|
| Command A Pro | Cohere | 200000 | $14.28 | 81.3 |
| Doubao 1.5 Pro Max | ByteDance | 400000 | $14.93 | 81.2 |
| Nova Pro Max | Amazon | 1000000 | $14.05 | 81.2 |
| Mistral Large 3 Instruct | Mistral AI | 400000 | $5.31 | 81.1 |
| Sonar Pro Max | Perplexity | 128000 | $14.01 | 81.1 |
| Claude 3.7 Sonnet Vision | Anthropic | 400000 | $13.96 | 81.0 |
FAQ
How much does Mistral Large 3 Instruct cost?
Mistral Large 3 Instruct is listed at $5.31 per million input tokens and $22.47 per million output tokens. Cached input is $0.63 per million tokens. A blended 3:1 input-to-output mix is about $9.60 per million tokens.
What is the context window of Mistral Large 3 Instruct?
Mistral Large 3 Instruct has a context window of 400,000 tokens and can emit up to 65,536 output tokens per request in this index.
Is Mistral Large 3 Instruct open weights?
Yes. Mistral Large 3 Instruct is listed as an open-weight model under a open weights licence.
How does Mistral Large 3 Instruct score on benchmarks?
Mistral Large 3 Instruct averages 81.1 across the scored benchmarks in this index, ranking #99 of 600 models. Its strongest relative result is Arena Elo at 1446.2.
What is Mistral Large 3 Instruct best for?
Mistral Large 3 Instruct 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 $5.31 per million input tokens, it sits in the mid-range of the index, and the 400,000-token context window is enough for large files and multi-file diffs.