Mistral Large 4 (Le Chonk) is Mistral AI's open weight multimodal Mixture of Experts model · about 1T total and 52B active parameters, 128 routed experts with 4 active per token · text and image input, text output · 1M token context · three reasoning modes (low, medium, high) · weights released 2026-10-29 on Hugging Face as mistralai/Mistral-Large-4-1T-A52B under the Mistral Ecosystem Open-Weight License, a modified MIT license that requires a commercial license for Model-as-a-Service use.
Tested on 24 benchmarks · BenchGecko score 39.3. Top scores: OTIS Mock AIME 2024-2025 (99.4%), GPQA diamond (90.5%), Artificial Analysis · Long Context Reasoning (81.3%).
ERNIE 4.5 21B A3B Thinking scores 39.8 (101% as good) at $0.07/1M input · 90% cheaper
Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.
Graduate-level science questions written by PhD experts. Diamond subset contains questions where experts disagree, testing deep understanding.
Tactical chess puzzles testing pattern recognition and multi-move calculation. Measures strategic reasoning ability.
Simple factual questions with verified correct answers. Tests accuracy of basic knowledge retrieval. Low scores indicate hallucination.
- Typemultimodal
- Context1.0M tokens (~524 books)
- ReleasedOct 2026
- LicenseProprietary
- StatusActive
- Cost / Message~$0.004
Frequently Asked Questions
Related Models
Claude 3.5 Sonnet · AnthropicQwen3.5-9B · Alibaba QwenGPT-4o (2024-08-06) · OpenAIBaichuan 2-7B · UnknownQwen3 14B · Alibaba QwenKey facts · as of 2026-11-12
- Mistral Large 4 by Mistral AI. BenchGecko score 39.3, rank 214 of 327 scored models (normalized average of public benchmark scores).
- List price $0.70 input · $2.10 output per 1M tokens (as of 2026-11-12).
- Sold by 2 providers (as of 2026-11-12): Mistral $0.70 in / $2.10 out · Mistral $0.77 in / $2.31 out. Every provider
- Gecko Tests · Gecko Score 54 (rank 25): Who Are You B (Knows who made it) · Censorship Index A (Answers almost everything) · Knowledge Horizon E (Knows news up to Apr 2025).
How to cite · data as of 2026-11-12
Mistral Large 4 · benchmarks, pricing and providers. BenchGecko, data as of 2026-11-12. https://benchgecko.ai/model/mistral-large-4-0
Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP