minimax
MiniMax: MiniMax M2.7
minimax/minimax-m2.7
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
MiniMax M2.7 is a text-to-text model listed on OpenRouter with a long context window, mandatory reasoning, and API controls for tools, structured responses, sampling, seeds, and stopping. MiniMax describes it as an agent-oriented model that can build complex harnesses, coordinate Agent Teams, use dynamic tool search, and handle software-engineering and productivity work. Its documentation reports strong benchmark outcomes across coding, ML, terminal, and office/tool-use tasks, but the source pack relies on catalog data and vendor claims, so independent verification and workload testing are still needed.[1][2]
Strengths
- OpenRouter's catalog marks reasoning as mandatory and lists text generation with a long context window, tool calls, tool choice, structured response formats, and common sampling controls.[1]
- MiniMax describes a development-time self-evolution loop in which the model updated its memory, built Skills for RL experiments, and iteratively improved a programming scaffold.[2]
- Vendor-reported measurements cover ML-competition, software-engineering, terminal, office-document, and tool-use benchmarks, including multi-agent collaboration.[2]
Limitations
- The pack contains catalog data and MiniMax's own documentation, but no independent model card or third-party evaluation; validate quality, latency, and cost on your workload.[1][2]
- The catalog lists only text input and output; multimodal support is not indicated.[1]
- The Hugging Face metadata labels the license as other without summarizing terms; review the referenced license before redistribution or deployment.[2]
Where to use it
- Evaluate it as the reasoning engine for tool-using coding agents that triage bugs, inspect logs, refactor code, and analyze production incidents.[1][2]
- Prototype multi-agent workflows where Agent Teams, skill use, and dynamic tool search may help coordinate complex tasks.[2]
- Consider it for long-context text pipelines such as repository, document, or ML-experiment analysis after confirming output quality and operational fit.[1][2]
Sources
- [1] OpenRouter catalog · MiniMax: MiniMax M2.7Sources checked: Nov 13, 2026
- [2] Model documentation · MiniMax: MiniMax M2.7Sources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Built to actively participate in its own evolution, M2.7 integrates advanced agentic capabilities through multi-agent...
Specifications
- Model ID
- minimax/minimax-m2.7
- Canonical ID
- minimax/minimax-m2.7-20260318
- Creator
- minimax
- Added to OpenRouter
- 18.03.2026
- Context length
- 196,608
- Modality
- text->text
- Input modalities
- text
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.210 / out $0.840
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | reasoning: mandatory · parameters: include_reasoning, reasoning |
| Coding | BB — next 30% of the cohort | — |
| Agents | BB — next 30% of the cohort | parameters: tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: mandatory · parameters: include_reasoning, reasoning |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Coding index | 52.6 | BB — next 30% of the cohort | 13.11.2026 |
| Artificial Analysis | Agentic index | 15.3 | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · 3dwin rate 49.9% · rank 57 | 1,178Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · asciiartwin rate 45.9% · rank 54 | 1,129Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · codecategorieswin rate 50.8% · rank 56 | 1,224Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · datavizwin rate 52.4% · rank 45 | 1,230Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · gamedevwin rate 49.1% · rank 56 | 1,199Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · svgwin rate 46.6% · rank 60 | 1,125Elo | CC — below the B band | 13.11.2026 |
| Design Arena | models · uicomponentwin rate 48.6% · rank 60 | 1,203Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · websitewin rate 51.3% · rank 53 | 1,235Elo | BB — next 30% of the cohort | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| GMICloudfp8 | 196,608 | $0.210 | $0.840 | — | — | 93.33% |
| Groq | 196,608 | $0.600 | $1.8 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.