MiniMax M2.7 Is Now Available in Puter.js

Puter.js now supports MiniMax M2.7, the latest proprietary model from MiniMax — and the first commercial model to actively participate in its own training through autonomous self-evolution loops.

What is MiniMax M2.7?

MiniMax M2.7 is a proprietary reasoning model released on March 18, 2026, that pushes the boundaries of agentic AI. What makes it unique is its self-evolving training approach — the model autonomously ran 100+ iterative loops of analyzing failure trajectories, modifying scaffold code, running evaluations, and deciding whether to keep or revert changes, achieving a 30% performance improvement on internal evals. Key highlights include:

• Self-Evolving AI: First commercial model to handle 30–50% of its own reinforcement learning research workflow autonomously
• Elite Agentic Coding: 56.2% on SWE-Pro and 55.6% on VIBE-Pro, approaching Claude Opus 4.6 levels
• Complex Agent Systems: 97% skill adherence rate across 40+ complex skills (>2000 tokens each), with support for Agent Teams and dynamic tool search
• Office Productivity: Significantly improved complex editing in Excel, PowerPoint, and Word with multi-round revisions
• Cost Efficient: At $0.30/$1.20 per million tokens (input/output), roughly 95% cheaper than comparable frontier models

Examples

Agentic coding workflow

puter.ai.chat("Analyze this production log file, identify the root cause of the memory leak, and generate a fix with proper error handling",
  { model: 'minimax/minimax-m2.7' }
);

Complex agent system

puter.ai.chat(`Build an agent workflow that:
1. Monitors a GitHub repo for new issues
2. Classifies issues by severity and component
3. Drafts initial responses with relevant code context
4. Creates linked pull requests for straightforward fixes`,
  { model: 'minimax/minimax-m2.7', stream: true }
);

Software engineering task

puter.ai.chat("Refactor this monolithic service into microservices with proper API boundaries, shared types, and integration tests",
  { model: 'minimax/minimax-m2.7' }
);

Office document generation

puter.ai.chat("Create a quarterly engineering report in a structured format with performance metrics, incident summaries, and team velocity charts",
  { model: 'minimax/minimax-m2.7' }
);

Get Started Now

Just add one library to your project:

// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';

Or add one script tag to your HTML:

<script src="https://js.puter.com/v2/"></script>

No API keys needed. Start building with MiniMax M2.7 immediately.

Learn more:

• Free, Unlimited MiniMax API
• Puter.js AI Documentation