priorsBuilder

Xiaomi · released 2026-09-21

MiMo V2.6 Pro

Measured on 56 benchmarks.

General frontier

49.4

#35 of 136 models

includes estimates: 24% of the mix is measured for this model

Your mix’s benchmarks

2 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #28default of 58749.4%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #95default of 5268.4

Leader: Opus 5.5 · 46.4

All 56 benchmarks

show 54 more

agentic

AA-AnalystAgent

44 results
  1. #22default of 4437.5%

Leader: Gemini 3.7 Flash · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #30default of 2241517

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #20high of 503.5%

Leader: Opus 5.5 · 15.5%

agentic

AutomationBench-AA

219 results
  1. #40default of 21958.6%

Leader: Gemini 4 · 77.5%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #17high of 7910.8%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #15high of 7847.1%

Leader: Muse Spark 1.3 · 55.3%

agentic

Tax Agent Bench

69 results
  1. #17high of 6932.4%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #17high of 4731.3%

Leader: Opus 5.5 · 65.2%

agentic

Terminal-Bench Science (Vals)

40 results
  1. #32high of 402.9%

Leader: GPT-6 Astra · 62.9%

coding

Arena Image-to-WebDev

58 results
  1. #17high of 581595

Leader: Opus 5.5 · 1749

coding

Arena WebDev

122 results
  1. #18high of 1221629

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #26high of 7743.0%

Leader: Sonnet 5.5 · 69.8%

coding

IOI

44 results
  1. #36high of 4439.3%

Leader: Gemini 4 · 100.0%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #25high of 6212.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #28high of 620.5%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #22high of 6269.8%

Leader: Opus 5.5 · 87.0%

coding

SciCode

226 results
  1. #9default of 22660.9%

Leader: Opus 5.5 · 66.9%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #33high of 7467.8%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Science-AA

46 results
  1. #32default of 465.7%

Leader: GPT-6 Astra · 63.3%

coding

Terminal-Bench v4-AA

216 results
  1. #34default of 21634.8%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #14high of 10585.2%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #33default of 61046.3

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #25high of 2201480

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #69high of 2201456

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #14high of 4755.2%

Leader: Gemini 4 · 68.9%

cost

AA Intelligence Index Cost per Task

177 results
  1. #47default of 177$0.13

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #42default of 177$207

Leader: GPT-6 Luna · $10.63

economics

Excel Modeling Benchmark

74 results
  1. #29high of 7462.9%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #14high of 7857.3%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #67default of 21019.2%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #14default of 2881682

Leader: Opus 5.5 · 1866

knowledge

AA-Omniscience Accuracy

526 results
  1. #152default of 52634.8%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #78default of 52659.4%

Leader: MiniCPM5-1B (Non-reasoning) · 99.1%

knowledge

MedCode

94 results
  1. #36high of 9445.0%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #12high of 9688.3%

Leader: Opus 5.5 · 91.4%

knowledge

Public Benefits Bench v1.1

50 results
  1. #8high of 5068.9%

Leader: Opus 5 · 76.9%

long-context

AA-LCR v1.1

523 results
  1. #3default of 52386.3%

Leader: Kimi K3 · 88.7%

long-context

MLCR-AA

94 results
  1. #34default of 9418.3%

Leader: Sonnet 5.5 · 75.0%

math

ProofBench v1.1

52 results
  1. #20high of 5270.0%

Leader: Sonnet 5.5 · 100.0%

other

Arena Vision

116 results
  1. #55high of 1161246

Leader: Fable 5 · 1308

reasoning

CritPt

529 results
  1. #37default of 52926.6%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Humanity's Sixth Sense

25 results
  1. #19high of 2525.2%

Leader: GPT-6 Astra · 53.6%

reasoning

MysteryMechanism

27 results
  1. #25high of 2715.3%

Leader: GPT-6 Astra · 53.2%

reasoning

Roboflow Visual Reasoning

102 results
  1. #72high of 10255.9%
  2. #95low of 10235.1%

Leader: GPT-6 Astra · 91.2%

reasoning

SAGE

80 results
  1. #41high of 8045.0%

Leader: Opus 4.7 · 56.1%

reliability

Tool-call error rate

171 results
  1. #154high of 1716.7%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #138high of 18198.4%

Leader: GPT-5 Pro · 100.0%

safety

CWE-Bench

19 results
  1. #5default of 1963.3%

Leader: Grok 4.7 · 68.3%

safety

CyberBench v1.1

45 results
  1. #15high of 4572.9%

Leader: Wonder · 80.4%

safety

SRE Bench

32 results
  1. #20high of 323.1%

Leader: GPT-6 Astra · 56.9%

speed

Latency (Time to First Token)

351 results
  1. #224default of 3513.9s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #300default of 35147.9

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #162high of 1774.0s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #159high of 17631.5

Leader: Muse Spark 1.1 · 204.3