Model comparison
Mistral Small 3 vs Llama 3.1
- Input
- $0.10 / 1M tokens
- Output
- $0.30 / 1M tokens
- Context
- 32k tokens
- Released
- 2025-01-30
- Input
- $2.75 / 1M tokens
- Output
- $6.50 / 1M tokens
- Context
- 128K
- Released
- 2024-07-23
Llama 3.1 scores higher on 5 of the 5 benchmarks both models report; Mistral Small 3 is cheaper on input tokens ($0.1 vs $2.8 per 1M, 27.5×); Llama 3.1 has the larger context window (128K vs 32K).
Specs and pricing side by side
| Spec | Mistral Small 3 | Llama 3.1 |
|---|---|---|
| Maker | Mistral AI | Meta |
| Released | 2025-01-30 | 2024-07-23 |
| Context window | 32k tokens | 128K |
| Max output | Not separately published by Mistral; bounded… | Provider-dependent (e.g. ~4K on some hosted… |
| Input price | $0.10 / 1M tokens | $2.75 / 1M tokens |
| Output price | $0.30 / 1M tokens | $6.50 / 1M tokens |
| Parameters | 24B | 8B, 70B, and 405B (dense) |
| License | Apache-2.0 | Llama 3.1 Community License (custom commercial) |
| Weights | Open weights | Open weightssame |
| Modalities | Text | Textsame |
| Knowledge cutoff | October 2023 | December 2023 |
Benchmarks both models report
| Benchmark | Mistral Small 3 | Llama 3.1 |
|---|---|---|
| MMLU | 81% | 87.3% |
| MMLU Pro | 66.3% | 73.3% |
| HumanEval | 84.8% | 89% |
| MATH | 70.6% | 73.8% |
| IFEval | 82.9% | 88.6% |
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 Small 3 if…
- Cost matters: input tokens cost 27.5× less ($0.1 vs $2.8 per 1M).
- Your use case is: fast-response conversational assistants and chatbots.
Pick Llama 3.1 if…
- You care about MMLU, MMLU Pro, HumanEval (it scores higher on 5 shared benchmarks).
- You need the longer context (128K vs 32K).
- Your use case is: self-hosted and on-prem deployments where data cannot leave the org's infrastructure.
FAQ
Which is better, Mistral Small 3 or Llama 3.1?
Llama 3.1 scores higher on 5 of the 5 benchmarks both models report; Mistral Small 3 is cheaper on input tokens ($0.1 vs $2.8 per 1M, 27.5×); Llama 3.1 has the larger context window (128K vs 32K). The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How much cheaper is Mistral Small 3?
Mistral Small 3: $0.10 / 1M tokens input, $0.30 / 1M tokens output. Llama 3.1: $2.75 / 1M tokens input, $6.50 / 1M tokens output (list API prices as published by each maker).
What are the context windows of Mistral Small 3 and Llama 3.1?
Mistral Small 3 offers 32k tokens with up to Not separately published by Mistral; bounded by the 32k context window output; Llama 3.1 offers 128K with up to Provider-dependent (e.g. ~4K on some hosted endpoints); the 128K window is shared between input and output output.
Is Mistral Small 3 newer than Llama 3.1?
Mistral Small 3 was released 2025-01-30; Llama 3.1 was released 2024-07-23.