MiniMax-M2.5
About MiniMax-M2.5
MiniMax-M2.5 is MiniMax's latest large language model, extensively trained with reinforcement learning across hundreds of thousands of complex real-world environments. Built on a 229B-parameter MoE architecture, it achieves SOTA performance in coding, agentic tool use, search, and office work, scoring 80.2% on SWE-Bench Verified with 37% faster inference than M2.1
Leverage MiniMax-M2.5's 229B MoE architecture and reinforcement learning optimizations to power high-performance, cost-effective agentic workflows.
Architect-Level System Design
Move beyond snippets to full-lifecycle development. M2.5 plans system architecture, UI design, and backend logic across multiple platforms.
Use Case Example:
"Designed a cross-platform Rust and Flutter real-time messaging app, generating the complete backend API schema and environment setup in a single session."
Autonomous Professional Research
Deploy agents for deep web exploration. M2.5 uses RISE-level search capabilities to synthesize information from dense, professional-grade sources.
Use Case Example:
"Conducted a deep-dive competitive analysis on solid-state battery patents, navigating 50+ technical papers to produce a synthesized market landscape report."
High-Value Office Automation
Automate complex deliverables in finance and law. M2.5 handles professional-grade Excel modeling, Word drafting, and PowerPoint generation.
Use Case Example:
"Built a dynamic Excel financial model for a Series B startup, integrating automated sensitivity analysis and professional-grade investor slide decks."
High-Throughput Code Auditing
Perform massive-scale codebase reviews at 100 tokens per second. Identify logical flaws and optimize performance with 'too cheap to meter' efficiency.
Use Case Example:
"Audited a 100,000-line Go microservices codebase for concurrency bugs, identifying three critical race conditions with 37% faster turnaround than previous models."
Metadata
Specification
State
Deprecated
Architecture
MoE
Calibrated
No
Mixture of Experts
Yes
Total Parameters
229B
Activated Parameters
Reasoning
No
Precision
FP8
Context length
197K
Max Tokens
Compare with Other Models
See how this model stacks up against others.
MiniMaxAI
chat
MiniMax-M2.1
Total Context:
197K
Max output:
131K
Input:
$
0.29
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M2
Total Context:
197K
Max output:
131K
Input:
$
0.3
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M1-80k
Total Context:
131K
Max output:
131K
Input:
$
0.55
/ M Tokens
Output:
$
2.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M3
Total Context:
1049K
Max output:
131K
Input:
$
0.3
/ M Tokens
Output:
$
1.2
/ M Tokens
MiniMaxAI
chat
MiniMax-M2.5
Total Context:
197K
Max output:
Input:
$
0.3
/ M Tokens
Output:
$
1.2
/ M Tokens
