Back to Models
MiniMaxAIProprietary

MiniMax-M2.7

MiniMax-M2.7 is a reasoning model developed by MiniMaxAI using MoE (Mixture of Experts) architecture. It supports an extensive context window of 200K tokens and features advanced self-improvement capabilities that drive the model's own training cycle.

Parameters

Undisclosed

Context Window

License

https://github.com/MiniMax-AI/MiniMax-M2.7/blob/main/LICENSE

Release Date

2026-03-18

API Pricing

Input Price (per 1M tokens)

$0.3

Output Price (per 1M tokens)

$

Billing Mode: standard

Strengths

  • High software engineering capability
  • Efficient MoE architecture
  • Strong self-improvement workflow

Weaknesses

  • Closed-source availability
  • Room to improve tool-calling accuracy
  • Limited API access environment

Use Cases

  • Advanced code generation and modification
  • Large-scale context analysis
  • Automation of specialized office tasks