MiniMax-M2
About MiniMax-M2
MiniMax-M2 redefines efficiency for agents. It's a compact, fast, and cost-effective MoE model (230 billion total parameters with 10 billion active parameters) built for elite performance in coding and agentic tasks, all while maintaining powerful general intelligence. With just 10 billion activated parameters, MiniMax-M2 provides the sophisticated, end-to-end tool use performance expected from today's leading models, but in a streamlined form factor that makes deployment and scaling easier than ever
Explore how MiniMax-M2's compact efficiency and powerful agentic intelligence excel in complex coding, debugging, and automated agent workflows.
Advanced Code Generation & Refactoring
Accelerate development with MiniMax-M2's advanced coding capabilities, handling multi-file edits, coding-run-fix loops, and test-validated repairs for efficient software delivery.
Example:
"Developed a new user authentication module for a Node.js application, automatically generating boilerplate, integrating with an existing database schema, and fixing integration tests."
Intelligent Agent Workflow Automation
Leverage MiniMax-M2's agentic performance to orchestrate complex, multi-step workflows, integrating shell, browser, retrieval, and custom code for seamless automation.
Example:
"Automated the deployment of a machine learning model, from fetching data via web scraping, training the model in a Python environment, to deploying it on a cloud instance using shell scripts."
Cross-Stack Debugging & Optimization
Utilize MiniMax-M2 to pinpoint subtle logical errors and suggest performance optimizations across diverse programming languages and complex technology stacks, enhancing code quality.
Example:
"Pinpointed a critical latency bottleneck in a microservices architecture, tracing the issue from a C# API gateway to a Python data processing service and proposing a caching strategy."
Automated Code & Security Audits
Deploy MiniMax-M2 for automated auditing of codebases and system configurations, reasoning through logical dependencies to identify inconsistencies, vulnerabilities, and compliance gaps.
Example:
"Audited a smart contract written in Solidity, identifying a reentrancy vulnerability and suggesting a secure pattern implementation, preventing potential financial loss."
Metadata
Specification
State
Deprecated
Architecture
Calibrated
No
Mixture of Experts
Yes
Total Parameters
230B
Activated Parameters
10B
Reasoning
No
Precision
FP8
Context length
197K
Max Tokens
131K
Compare with Other Models
See how this model stacks up against others.
MiniMaxAI
chat
MiniMax-M2.5
Release on: Feb 15, 2026
Total Context:
197K
Max output:
131K
Input:
$
0.3
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M2.1
Release on: Dec 23, 2025
Total Context:
197K
Max output:
131K
Input:
$
0.29
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M2
Release on: Oct 28, 2025
Total Context:
197K
Max output:
131K
Input:
$
0.3
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M1-80k
Release on: Jun 17, 2025
Total Context:
131K
Max output:
131K
Input:
$
0.55
/ M Tokens
Output:
$
2.2
/ M Tokens
