MiniMax M2.7
MiniMax M2.7 from MiniMax is built for complex agent teams, self-evolution, software engineering, and office editing with strong tool use and workflows.
Model Overview
Capabilities, design details, and architectural traits
MiniMax M2.7 - self-evolving agent model
MiniMax M2.7 is built around self-improvement and complex agent workflows. It is presented as a model for software engineering, office tasks, and multi-agent coordination.
| Trait | Documented detail |
|---|---|
| Self-improvement loop | It is described as a model that participated in its own evolution and uses repeated optimization cycles. |
| Agent Teams | It is built to work with multiple agents and to handle complex agent harnesses. |
| Tool use | It supports tool interaction and dynamic tool search during task execution. |
| Office editing | It is documented for multi-turn, high-fidelity editing in Excel, PowerPoint, and Word. |
| Coding and engineering | It is highlighted for end-to-end software delivery, bug hunting, security analysis, and ML tasks. |
Agent workflow
The model is framed as a system for handling hard, multi-step productivity work. Its identity is tied to repeated refinement, tool use, and coordinated execution across agent roles.
Workspace focus
It is also positioned for professional office work, with stronger handling of complex edits and document workflows. The model page emphasizes productivity tasks rather than general chat alone.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | MiniMax | Anthropic | Anthropic |
| Release Date | March 18, 2026 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | - | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 205K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.30 Best Input Pricing | $4 | $2 |
| Output Pricing | $1.20 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | text | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 22.8 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | 52.6 | - | - |
| Agentic Index | 15.3 | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from MiniMax
Other models by MiniMax
Top AI Models
Leading alternatives by intelligence score
