Bitcoin.com AI logo

MiMo V2.6 Pro

MiMo V2.6 Pro is Xiaomi's flagship native omnimodal model, built with 1.02T total parameters and 42B active parameters per token. It is designed for demanding coding, long-horizon agent workflows, visual tasks, research, and cybersecurity, with text, image, video, and audio understanding and a 1M-token context window.

  • Reasoning
  • Vision
  • Audio Input
  • Video Input
  • Tool Calling
  • Structured Output

Added Sep 21, 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)
1.2K tokens
Longer than 78% of models

Benchmarks

Sourced from Artificial Analysis.

Intelligence Index

46.3

Better than 96% of models compared

Agentic work

AutomationBench-AA

Workflow automation with guardrail penalties

58.6%

Better than 82% of models compared

AutomationBench-AA Tasks Completed

Fully completed workflows without guardrail violations

25.4%

Better than 35% of models compared

AA-Briefcase

Agentic knowledge work (Elo)

1516 Elo

Better than 88% of models compared

GDPval-AA v2

Economically valuable tasks (Elo)

1686 Elo

Better than 96% of models compared

Document reasoning

GDP.pdf

Professional PDF reasoning: all-pass rate

19.2%

Better than 68% of models compared

AA-LCR v1.1

Long context reasoning with updated grading

86.3%

Better than 99% of models compared

MLCR-AA

Medical long-context reasoning

18.3%

Better than 63% of models compared

Reasoning

HLE

Humanity's Last Exam

49.4%

Better than 96% of models compared

CritPt

Research-level physics reasoning

26.6%

Coding

Terminal-Bench v4.0

Practical coding and terminal tasks

34.8%

Better than 85% of models compared

SciCode

Python programming for scientific computing

60.9%

Better than 96% of models compared

Knowledge

AA-Omniscience Accuracy

Proportion of correctly answered questions

34.8%

AA-Omniscience Hallucination Rate

Rate of incorrect answers among non-correct responses

40.6%

Legacy benchmarks

AA-LCR (unversioned / legacy)

Long context reasoning evaluation

86.3%

Better than 99% of models compared

GDPval-AA (unversioned / legacy)

Economically valuable tasks

59.2%

Last updated Oct 14, 2026

Artificial Analysis

Providers

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

Loading provider options…