MiMo V2.5 Pro Thinking

MiMo V2.5 Pro with Xiaomi thinking enabled for coding, long-context reasoning, and agentic orchestration.

Reasoning
Tool Calling
Structured Output

Added Jun 3, 2026

Model weights

Pricing

Auto routing · per 1M tokens

Input: $0.43
Output: $0.87
Cache read: $0.0036

Compare provider prices

Specifications

Context window: 1M
Max output: 131.1K
Parameters: 1T / 42B; Total / active
Avg output (7d): 1.3K tokens; Longer than 79% of models

Benchmarks

Sourced from Artificial Analysis.

Intelligence Index

26.0

Better than 79% of models compared

Coding Index

60.2

Better than 72% of models compared

Agentic Index

21.3

Better than 59% of models compared

Agentic work

AutomationBench-AA

Workflow automation with guardrail penalties

13.7%

Better than 35% of models compared

AutomationBench-AA Tasks Completed

Fully completed workflows without guardrail violations

2.0%

Better than 36% of models compared

AA-Briefcase v1.1

Agentic knowledge work (Elo)

883 Elo

Better than 37% of models compared

GDPval-AA v2.1

Economically valuable tasks (Elo)

1121 Elo

Better than 54% of models compared

Document reasoning

GDP.pdf

Professional PDF reasoning: all-pass rate

4.0%

Better than 23% of models compared

AA-LCR v1.1

Long context reasoning with updated grading

79.7%

Better than 84% of models compared

MLCR-AA

Medical long-context reasoning

5.6%

Better than 17% of models compared

Reasoning

HLE

Humanity's Last Exam

35.7%

Better than 83% of models compared

IFBench

Instruction-following benchmark

79.9%

Better than 98% of models compared

CritPt

Research-level physics reasoning

4.0%

Coding

Terminal-Bench v4.0

Practical coding and terminal tasks

0.0%

Better than 16% of models compared

SciCode

Python programming for scientific computing

50.6%

Better than 56% of models compared

Knowledge

AA-Omniscience Accuracy

Proportion of correctly answered questions

22.4%

AA-Omniscience Hallucination Rate

Rate of incorrect answers among non-correct responses

24.7%

Legacy benchmarks

GPQA Diamond (legacy)

Graduate-level scientific reasoning

86.6%

Better than 81% of models compared

Terminal-Bench Hard (legacy)

Agentic coding and terminal use

43.2%

Better than 90% of models compared

T²-Bench Telecom (legacy)

Conversational AI agents in dual-control scenarios

94.2%

Better than 92% of models compared

AA-LCR (unversioned / legacy)

Long context reasoning evaluation

79.7%

Better than 84% of models compared

GDPval-AA (unversioned / legacy)

Economically valuable tasks

31.0%

Last updated Oct 22, 2026

Artificial Analysis

Providers

Choose explicit providers for this model. Auto routing remains available as the default option.

Loading provider options…