SIDE BY SIDE
Mistral Large 3 2512 vs Mistral Medium 3.1
Both are Mistral models. Mistral Large 3 2512 was added to OpenRouter on 2025-12-01, Mistral Medium 3.1 on 2025-08-13.
| Field | Mistral Large 3 2512 | Mistral Medium 3.1 |
|---|---|---|
| Provider | Mistral | Mistral |
| Family | Mistral | Mistral |
| Province | Maple Highlands | Maple Highlands |
| Context window | 262K tokens | 131K tokens |
| Maximum output | 210K tokens | 105K tokens |
| Input / 1M tokens | $0.500 | $0.400 |
| Output / 1M tokens | $1.50 | $2.00 |
| Accepts | text, files, images | text, files, images |
| Produces | text | text |
| Reasoning | No | No |
| Tool use | Yes | Yes |
| Added to OpenRouter | 2025-12-01 | 2025-08-13 |
| Model ID | mistralai/mistral-large-2512 | mistralai/mistral-medium-3.1 |
Questions and answers
Which is cheaper, Mistral Large 3 2512 or Mistral Medium 3.1?
It depends on the mix: Mistral Medium 3.1 is cheaper for input ($0.400 vs $0.500), Mistral Large 3 2512 for output ($1.50 vs $2.00) per million tokens. At three input tokens per output token, Mistral Large 3 2512 costs less. On input, Mistral Medium 3.1 is 20% cheaper.
Which has the larger context window?
Mistral Large 3 2512: 262K tokens against 131K for Mistral Medium 3.1 (2× as long).
What can one do that the other cannot?
Nothing among reasoning, tool use and image input: both support tool use and image input.
Answers are computed from the 2026-10-24 catalog snapshot.