MiniMax: minimax-m2.1 API
MiniMax-M2.1 is a major evolution of the M2 architecture, released in late 2025. Its primary breakthrough lies in Multilingual Programming SOTA performance, excelling not only in Python but also in Rust, Go, Swift, and Kotlin. Retaining the 230B/10B MoE structure, M2.1 introduces a refined "Concise CoT" mechanism, delivering significantly faster response times and reducing token consumption by 30%. Designed as a high-performance brain for autonomous agents, it excels in full-stack development, from backend architecture to native iOS/Android app delivery.
- Context window: 204,800 tokens
- Max output: 131,072 tokens
- Input: text
- Output: text
- Reasoning: Supported
- Tool calling: Supported
- Released: 2025-12-23
- Knowledge cutoff: 2025-05-31
Frequently Asked Questions
What is the context window of minimax-m2.1?
minimax-m2.1 supports a context window of up to 204,800 tokens.
Does minimax-m2.1 support function calling?
Yes. minimax-m2.1 supports tool / function calling.
Does minimax-m2.1 support reasoning?
Yes. minimax-m2.1 is a reasoning-capable model.
What is the knowledge cutoff of minimax-m2.1?
The knowledge cutoff of minimax-m2.1 is 2025-05-31.