Model comparison
Mistral Medium 3 vs Llama 4 Scout
- Input
- $0.40 / 1M tokens
- Output
- $2.00 / 1M tokens
- Context
- 128K
- Released
- 2025-05-07
- Input
- $0.18
- Output
- $0.59
- Context
- 10M
- Released
- 2025-04-05
Llama 4 Scout scores higher on 2 of the 3 benchmarks both models report (the other leads on 1); Llama 4 Scout is cheaper on input tokens ($0.2 vs $0.4 per 1M, 2.2×); Llama 4 Scout has the larger context window (10M vs 128K); Llama 4 Scout ships open weights while Mistral Medium 3 is API-only.
Specs and pricing side by side
| Spec | Mistral Medium 3 | Llama 4 Scout |
|---|---|---|
| Maker | Mistral AI | Meta |
| Released | 2025-05-07 | 2025-04-05 |
| Context window | 128K | 10M |
| Input price | $0.40 / 1M tokens | $0.18 |
| Output price | $2.00 / 1M tokens | $0.59 |
| Parameters | not published | 109B total · 17B active (16 experts) |
| License | Proprietary | Llama 4 Community License Agreement |
| Weights | API only | Open weights |
| Modalities | Text, Vision, PDF | Text, Vision |
| Knowledge cutoff | Not disclosed | 2024-08 |
Not published by either maker: max output.
Benchmarks both models report
| Benchmark | Mistral Medium 3 | Llama 4 Scout |
|---|---|---|
| MMLU Pro | 77.2% | 74.3% |
| GPQA Diamond | 57.1% | 57.2% |
| MMMU | 66.1% | 73.4% |
Scores are as published by each maker or its cited source; harness, reasoning effort and date can differ between rows, so treat small gaps as ties. Sources are listed on each model page.
Which one should you pick?
Pick Mistral Medium 3 if…
- You care about MMLU Pro (it scores higher on 1 shared benchmark).
- You need PDF input, which Llama 4 Scout lacks.
- Your use case is: cost-sensitive enterprise coding assistants and code generation.
Pick Llama 4 Scout if…
- You care about GPQA Diamond, MMMU (it scores higher on 2 shared benchmarks).
- Cost matters: input tokens cost 2.2× less ($0.2 vs $0.4 per 1M).
- You need the longer context (10M vs 128K).
- You want to self-host or fine-tune (open weights, Llama 4 Community License Agreement).
- Your use case is: long-document and large-codebase analysis at up to 10M-token context.
FAQ
Which is better, Mistral Medium 3 or Llama 4 Scout?
Llama 4 Scout scores higher on 2 of the 3 benchmarks both models report (the other leads on 1); Llama 4 Scout is cheaper on input tokens ($0.2 vs $0.4 per 1M, 2.2×); Llama 4 Scout has the larger context window (10M vs 128K); Llama 4 Scout ships open weights while Mistral Medium 3 is API-only. The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How much cheaper is Llama 4 Scout?
Mistral Medium 3: $0.40 / 1M tokens input, $2.00 / 1M tokens output. Llama 4 Scout: $0.18 input, $0.59 output (list API prices as published by each maker).
What are the context windows of Mistral Medium 3 and Llama 4 Scout?
Mistral Medium 3 offers 128K; Llama 4 Scout offers 10M.
Is Mistral Medium 3 newer than Llama 4 Scout?
Mistral Medium 3 was released 2025-05-07; Llama 4 Scout was released 2025-04-05.
Go deeper
- Mistral Medium 3: full specs, pricing and benchmarks
- Llama 4 Scout: full specs, pricing and benchmarks
- mistral medium 3 5 vs mistral medium 3
- mistral medium 3 vs mistral medium legacy
- llama 4 scout vs gemini 2 0 pro
- llama 4 scout vs deepseek v3
- mistral medium 3 vs gemma 3
- llama 4 scout vs qwen2 5 max
- All model comparisons