MMistral · Released Dec 1, 2025
Mistral Large 3 2512
Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.
- Released
- Dec 1 2025
- Context window
- 262K 210K out
- Input / 1M tokens
- $0.50
- Output / 1M tokens
- $1.50
Where to run it
16 providers compared
Cheapest: Mistral, Kilo Gateway, Vercel AI Gateway +10 more at $0.50 / $1.50 per 1M input / output tokens — 55% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| Mistralzdr | $0.50 | $1.50 | $0.05 | 262K | 210K | 100%24h | Cheapest |
| Kilo Gateway | $0.50 | $1.50 | $0.05 | 262K | 210K | — | Cheapest |
| Vercel AI Gatewayno tool calling | $0.50 | $1.50 | $0.05 | 262K | 256K | — | Cheapest |
| LLM Gatewayno tool calling | $0.50 | $1.50 | — | 262K | 262K | — | Cheapest |
| Pioneer | $0.50 | $1.50 | $0.50 | 256K | 131K | — | Cheapest |
| Amazon Bedrock | $0.50 | $1.50 | — | 262K | 8K | — | Cheapest |
| Prime Inference | $0.50 | $1.50 | — | 262K | 210K | — | Cheapest |
| Tempr Gateway | $0.50 | $1.50 | $0.05 | 262K | 262K | — | Cheapest |
| DevPass (LLM Gateway) | $0.50 | $1.50 | — | 262K | 262K | — | Cheapest |
| Ollama Cloud | $0.50 | $1.50 | — | 262K | 262K | — | Cheapest |
| Merge Gateway | $0.50 | $1.50 | $0.05 | 256K | 256K | — | Cheapest |
| Eden AI | $0.50 | $1.50 | $0.05 | 262K | 262K | — | Cheapest |
| Opper | $0.50 | $1.50 | — | 256K | 8K | — | Cheapest |
| LowRouter | $0.52 | $1.57 | $0.052 | 262K | 262K | — | +5% |
| Mistraleu | $0.55 | $1.65 | $0.055 | 262K | 210K | 100%24h | +10% |
| Cortecs | $0.61 | $1.84 | $0.061 | 256K | 256K | — | +23% |
| 302.AI | $1.10 | $3.30 | — | 128K | 262K | — | +120% |
Value
Price vs. performance
Mistral Large 3 2512 Other Mistral Other labs Value frontier · GPQA Diamond vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
8 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| SciCodeSoftware engineering · Research-coding problems from the natural sciences | 36.6% | #86of 102 | Claude Opus 5.566.9% |
| Lech Mazur WritingWriting & content · Short stories weaving 10 required elements, judged 0–10 | 6.90/10 | #21of 27 | GPT-58.60/10 |
| WebDev ArenaDesign & visual · People vote on which model builds the better website (Elo) | 1230 | #86of 90 | Claude Opus 5.51820 |
| GPQA DiamondReasoning & knowledge · PhD-level biology, chemistry and physics questions | 51.3% | #97of 122 | Gemini 4 Argon96.4% |
| SimpleBenchReasoning & knowledge · Common-sense trick questions about space, time and people | 22.5% | #48of 58 | Claude Fable 581.9% |
| CritPtReasoning & knowledge · Open-ended, research-level physics problems | 0.0% | #92of 111 | GPT-5.6 Sol32.3% |
| LMCAReasoning & knowledge · Judge the quality of conceptual arguments like experts | 16.7% | #101of 123 | Claude Sonnet 5.568.3% |
| OTIS Mock AIMEMaths · Competition-style problems harder than MATH Level 5 | 8.5% | #101of 120 | Gemini 4 Argon100.0% |
Specs
Details
- Model ID
mistralai/mistral-large-2512- Developer
- Mistral
- Release date
- Dec 1, 2025 ·
- Context window
- 262,144 tokens
- Max output
- 209,715 tokens
- API price
- $0.50 / $1.50 per 1M input / output tokens
- Providers
- 16
- Input types
- text, image, file
- Reasoning
- No
- Tool calling
- Yes
- Open weights
- No
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.