Mimo V2.5

LLM
Xiaomi

Mimo V2.5 is an agentic model from Xiaomi built for complex software engineering and long-horizon tasks. It supports up to 1M token context, handles thousands of tool calls autonomously, and scores highly on SWE-bench Pro.

Context tokens

1,000,000

Output tokens

131,072

Released

May 29, 2026

Capabilities

Tool use
Structured output
Thinking

Supported tools

No tools enabled.

Pricing

TypeCreditsUnits
Input3.99Credits per 1k tokens
Output18.62Credits per 1k tokens
Cache hit0.74Credits per 1k tokens

Variants

No variants available for this model.

Try This Model

Write a prompt and experiment with Mimo V2.5 in the model experiments page. You can compare it with other models side by side.