Minimax M2.7
Minimax M2.7 is MiniMax's high-capability agentic model targeting end-to-end software engineering: project delivery, log analysis, bug troubleshooting, and code security. It supports a context window of 204.8K tokens and a max output of 131.1K tokens per request.
import { streamText } from 'ai'
const result = streamText({ model: 'minimax/minimax-m2.7', prompt: 'Why is the sky blue?'})Playground
Try out Minimax M2.7 by MiniMax. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
| Provider |
|---|
| 197K | 0.3s | 26tps | $0.30/M | $1.20/M | — | — | 03/18/2026 | |||||
| 196K | 1.1s | 150tps | $0.30/M | $1.20/M | Read:$0.06/M Write:— | — | — | 03/18/2026 | ||||
| 205K | 3.3s | 31tps | $0.30/M | $1.20/M | Read:$0.06/M Write:$0.38/M | — | — | 03/18/2026 | ||||
| 205K | 1.6s | 55tps | $0.30/M | $1.20/M | Read:$0.06/M Write:— | — | — | 03/18/2026 |
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
More models by MiniMax
| Model |
|---|
| 205K | 1.1s | 49tps | $0.60/M | $2.40/M | Read:$0.06/M Write:$0.38/M | — | — |  | 03/18/2026 | ||||
| 1M | 0.5s | 205tps | $0.27/M | $0.95/M | Read:$0.03/M Write:$0.38/M | — | — |   +2 | 02/12/2026 | ||||
| 205K | 0.8s | 47tps | $0.60/M | $2.40/M | Read:$0.03/M Write:$0.38/M | — | — |  | 02/12/2026 | ||||
| 205K | 0.5s | 177tps | $0.30/M | $1.20/M | Read:$0.03/M Write:$0.38/M | — | — | | 10/27/2025 | ||||
| 205K | 0.8s | 70tps | $0.30/M | $1.20/M | Read:$0.03/M Write:$0.38/M | — | — | | 10/27/2025 | ||||
| 205K | 1.0s | 50tps | $0.30/M | $2.40/M | Read:$0.03/M Write:$0.38/M | — | — | | 10/27/2025 |
About Minimax M2.7
Released on March 18, 2026, Minimax M2.7 introduces native support for multi-agent collaboration and complex skill orchestration. Where earlier generations focused on single-agent task completion, Minimax M2.7 coordinates across agent networks, passing context, managing handoffs, and tracking dependencies between parallel workstreams.
The model also introduces dynamic tool search, letting agents discover and invoke relevant tools at runtime rather than relying on a pre-specified list. This expands the model's effective action space and makes it more adaptable to novel task types during long-horizon workflows.
Beyond software engineering, Minimax M2.7 handles professional office tasks. This makes it suitable for enterprise automation spanning document processing, financial modeling, and multi-system data workflows alongside its coding capabilities.
What To Consider When Choosing a Provider
- Configuration: For teams building multi-agent systems, Minimax M2.7's native collaboration capabilities reduce the custom orchestration code you need to coordinate agents.
- Zero Data Retention: AI Gateway supports Zero Data Retention for this model via direct gateway requests (BYOK is not included). To configure this, check the documentation.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use Minimax M2.7
Best For
- Multi-agent systems: Native agent-to-agent coordination without custom middleware
- Production debugging workflows: Investigation spanning multiple files, services, or repositories
- End-to-end project delivery: Full workflows from specification through implementation and deployment
- Enterprise office automation: Workflows combining software engineering with document and data processing
- Dynamic tool discovery: Pipelines that need runtime tool selection rather than fixed tool lists
Consider Alternatives When
- Single-agent workload: Your workflow doesn't benefit from orchestration features
- Critical response latency: Latency-sensitive interactive use where M2.7-highspeed is a better fit
- Lower cost sufficient: Earlier M2 generations cover your requirements at lower cost
Conclusion
Minimax M2.7 is the right choice when your application outgrows single-agent patterns and needs a model that can coordinate, delegate, and dynamically discover tools across a production workflow.
Frequently Asked Questions
What is native multi-agent collaboration in Minimax M2.7?
Minimax M2.7 operates within multi-agent networks, handling context passing, handoffs, and dependency tracking between agents without custom orchestration middleware.
What is dynamic tool search?
Instead of using a fixed tool list, Minimax M2.7 discovers and invokes relevant tools at runtime based on the task at hand. This expands its adaptability during long-horizon workflows.
How does Minimax M2.7 differ from M2.5?
M2.5 introduced planning-before-coding for single-agent workflows. Minimax M2.7 adds multi-agent coordination, complex skill orchestration, and dynamic tool search for production-grade distributed agent systems.
Does Minimax M2.7 support professional office tasks beyond coding?
Yes. It performs well on professional office tasks including document processing and data workflows alongside software engineering.
Is there a speed-optimized variant of Minimax M2.7?
Yes.
minimax/minimax-m2.7-highspeedis the throughput-optimized variant at roughly double the listed input and output rates versus the standard variant.How do I access Minimax M2.7 through the AI SDK?
Set the model identifier to
minimax/minimax-m2.7in your SDK configuration.Can Minimax M2.7 handle production debugging across multiple services?
Yes. Production debugging and end-to-end project delivery are key improvements in the Minimax M2.7 generation.