Llama 4 Maverick Instruct vs Mistral Large 3: which model should you choose?
We compared Llama 4 Maverick Instruct and Mistral Large 3 on price, context window and 12 benchmarks. Llama 4 Maverick Instruct and Mistral Large 3 are effectively tied on the index average; Llama 4 Maverick Instruct is cheaper on a blended 3:1 basis.
Llama 4 Maverick Instruct
Llama 4 Maverick Instruct is a multimodal model from Meta AI. It offers a 200,000-token context window, costs $5.33 per million input tokens and $23.04 per million output tokens, and averages 81.7 across the index's benchmark suite.
- Input
- $5.33 / 1M
- Output
- $23.04 / 1M
- Context
- 200,000
- Avg score
- 81.7
- Open weights
- Yes
Mistral Large 3
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.
- Input
- $5.41 / 1M
- Output
- $23.83 / 1M
- Context
- 128,000
- Avg score
- 81.7
- Open weights
- Yes
Specification comparison
| Spec | Llama 4 Maverick Instruct | Mistral Large 3 | Edge |
|---|---|---|---|
| Input price / 1M | $5.33 | $5.41 | A |
| Output price / 1M | $23.04 | $23.83 | A |
| Context window | 200,000 | 128,000 | A |
| Max output | 16,384 | 4,096 | A |
| Average score | 81.7 | 81.7 | Tie |
| Arena Elo | 1438.5 | 1448.6 | B |
| Open weights | Yes | Yes | Tie |
Benchmark comparison
Llama 4 Maverick Instruct wins 5, Mistral Large 3 wins 6, and 1 benchmarks are tied. The table shows the modelled score for each model and the gap.
| Benchmark | Llama 4 Maverick Instruct | Mistral Large 3 | Gap | Winner |
|---|---|---|---|---|
| MMLU-Pro | 84.6 | 84.6 | 0.0 | Tie |
| GPQA Diamond | 79.5 | 79.6 | 0.1 | Mistral Large 3 |
| SWE-bench Verified | 69.1 | 71.3 | 2.2 | Mistral Large 3 |
| HumanEval | 85.0 | 88.6 | 3.6 | Mistral Large 3 |
| MATH-500 | 83.4 | 84.1 | 0.7 | Mistral Large 3 |
| AIME 2025 | 79.9 | 78.3 | 1.6 | Llama 4 Maverick Instruct |
| LiveCodeBench | 75.8 | 74.4 | 1.4 | Llama 4 Maverick Instruct |
| IFEval | 88.9 | 89.8 | 0.9 | Mistral Large 3 |
| TruthfulQA | 84.0 | 83.7 | 0.3 | Llama 4 Maverick Instruct |
| HellaSwag | 89.6 | 88.2 | 1.4 | Llama 4 Maverick Instruct |
| MMMU | 79.0 | 76.2 | 2.8 | Llama 4 Maverick Instruct |
| Arena Elo | 1438.5 | 1448.6 | 10.1 | Mistral Large 3 |
Price analysis
On a blended 3:1 input-to-output basis, Llama 4 Maverick Instruct costs about $9.76 per million tokens and Mistral Large 3 costs about $10.02. Llama 4 Maverick Instruct is roughly 3% cheaper than Mistral Large 3. At high volume, that gap compounds quickly, so price should carry real weight unless the quality difference is decisive.
Llama 4 Maverick Instruct has the larger context window at 200,000 tokens, compared with 128,000 for Mistral Large 3. That matters for long documents, large codebases and multi-step agent traces.
Verdict
Across 12 shared benchmarks, Llama 4 Maverick Instruct wins 5, Mistral Large 3 wins 6, and 1 are tied. The two models are effectively tied on the index average at 81.7.
The price picture favours Llama 4 Maverick Instruct. The context picture favours Llama 4 Maverick Instruct. Those two dimensions often matter more than a few benchmark points once a model is good enough for the task.
If you need a single default, start with Llama 4 Maverick Instruct; it is cheaper on a blended basis and has the larger context window. Test it on your own traffic before committing.
Choose Llama 4 Maverick Instruct if
- Unit cost matters and you want the cheaper blended price.
- You need the larger 200,000-token context window.
- You need the larger 16,384-token output limit.
Choose Mistral Large 3 if
- You prefer Mistral AI's tooling or already use it.
FAQ
Is Llama 4 Maverick Instruct better than Mistral Large 3?
The two models are effectively tied on the index average at 81.7. The better choice depends on price, context and your task. Llama 4 Maverick Instruct wins 5 benchmarks, Mistral Large 3 wins 6.
Which is cheaper, Llama 4 Maverick Instruct or Mistral Large 3?
Llama 4 Maverick Instruct is cheaper on a blended 3:1 basis at about $9.76 per million tokens, versus $10.02 for Mistral Large 3.
Which has the larger context window?
Llama 4 Maverick Instruct has the larger context window at 200,000 tokens. Mistral Large 3 has 128,000 tokens.
Are these models open weights?
Llama 4 Maverick Instruct is open weights; Mistral Large 3 is open weights.
Which should I choose for production?
If you need a single default, start with Llama 4 Maverick Instruct; it is cheaper on a blended basis and has the larger context window. Test it on your own traffic before committing.