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.
Model catalog
Models
Every model in the catalog with source-linked pricing, context limits, provider availability, and published benchmark results.
Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use
5 providers
| Model | Creator | Score | Inputs | Context | Input | Output | Released | Compare |
|---|---|---|---|---|---|---|---|---|
| Mistral: Mistral Large 3 2512 (batch)mistralai/mistral-large-2512:batch | Mistral AI |
1069.0 | 262.144K | $0.25 | $0.75 | — | ||
| Trinity Large Thinkingarcee-ai/trinity-large-thinking | 1041.0 | 524.288K | $0.25 | $0.8 | 2026-04-01 |