Minimax M1 40K
Minimax develops the Minimax M 1 40K, a chat model powered by a hybrid Mixture-of-Experts (MoE) architecture combined with a lightning attention mechanism, allowing for efficient scaling of test-time compute. It is particularly suitable for complex tasks that require processing long inputs, with a context window of 10,240,000 tokens and native support for a context length of 1 million tokens.
Input
Output
Context
10240K
Max Output
40K
Parameters
456.1B
Technical Specifications
Model TypeChat
Context Window10,240,000 tokens
Max Output Tokens40,000 tokens
Parameters456.1B
Release DateJun 5, 2025
Training CutoffNot available
Licenseapache-2.0
Open SourceYes
Input Modalities
Text
Output Modalities
Text
Capabilities
Benchmarks
Artificial Analysis
7.5%
HLE
81.3%
AIME
68.2%
GPQA
37.8%
SciCode
97.2%
MATH 500
80.8%
MMLU Pro
13.7
Math
14.1
Coding
65.7%
LiveCodeBench
14.4
Intelligence
Estimated GPU Requirements for Minimax M1 40K
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Jun 23, 2026
Start building with the right model.
Automatically route workloads to the right model for every task, every time.