M2.5 Highspeed

LLM

MiniMax

M2.5 Highspeed is designed for fast, cost-efficient agentic and coding tasks, offering strong tool use, planning, and long-chain task execution with a 1M-token context window.

Context tokens

1,000,192

Output tokens

8,192

Docs

Model documentation

Released

Mar 2, 2026

Schema

Schema documentation

Capabilities

Tool use

Structured output

Multilingual

Supported languages

Supported tools

Seclai Web Tools
Fetch web pages and search the web from within agent prompt calls. Includes seclai_web_fetch for retrieving page content in markdown, HTML, or plain text, and seclai_web_search for finding relevant pages with content snippets.

Pricing

Type	Credits	Units
Input	7.98	Credits per 1k tokens
Output	31.92	Credits per 1k tokens
Cache hit	0.40	Credits per 1k tokens

Variants