priorsBuilder

Xiaomi · released 2026-04-22

MiMo V2.5 Pro

Measured on 57 benchmarks across 3 reasoning efforts.

General frontier

53.1

#31 of 136 models · best at xhigh

includes estimates: 21% of the mix is measured for this model

Reasoning efforts in your mix

xhigh53.1#67 of 313
default37.6#167 of 313
none24.8#250 of 313

Your mix’s benchmarks

3 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #112default of 58735.7%
  2. #248none of 58714.8%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #113default of 5263.3
  2. #305none of 526-37.8

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #46xhigh of 19662.0%
  2. #103none of 19638.0%

Leader: Opus 4.8 · 95.0%

All 57 benchmarks

show 54 more

agentic

AA-AnalystAgent

44 results
  1. #29default of 4420.0%

Leader: Gemini 3.7 Flash · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #138default of 224883

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #45xhigh of 50-7.8%

Leader: Opus 5.5 · 15.5%

agentic

APEX-Agents-AA

33 results
  1. #31default of 332.4%

Leader: Gemini 3.5 Flash · 47.1%

agentic

AutomationBench-AA

219 results
  1. #141default of 21913.7%

Leader: Gemini 4 · 77.5%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #47xhigh of 792.1%

Leader: Muse Spark 1.2 · 25.4%

agentic

ITBench-AA

45 results
  1. #24default of 4538.2%

Leader: GPT-5.6 Sol · 56.2%

agentic

Legal Research Bench

78 results
  1. #58xhigh of 7815.9%

Leader: Muse Spark 1.3 · 55.3%

agentic

Tax Agent Bench

69 results
  1. #54xhigh of 6911.5%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #45xhigh of 470.5%

Leader: Opus 5.5 · 65.2%

agentic

τ²-Bench Telecom

393 results
  1. #34default of 39394.2%
  2. #151none of 39372.5%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #145default of 2069.9%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena WebDev

122 results
  1. #61xhigh of 1221479

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #50xhigh of 7721.6%

Leader: Sonnet 5.5 · 69.8%

coding

DeepSWE v1

29 results
  1. #20xhigh of 2919.5%

Leader: GPT-5.5 · 70.0%

coding

LiveCodeBench (Vals)

127 results
  1. #64xhigh of 12781.4%

Leader: Fable 5.1 · 90.5%

coding

SciCode

226 results
  1. #100default of 22650.6%

Leader: Opus 5.5 · 66.9%

coding

SWE-bench Verified (Vals)

85 results
  1. #47xhigh of 8574.0%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #45xhigh of 7457.3%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Hard-AA

387 results
  1. #45default of 38743.2%
  2. #81none of 38735.6%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #83default of 23865.2%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #151default of 2160.0%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #62xhigh of 10534.1%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #152default of 61026.0
  2. #248none of 61018.3

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #44xhigh of 2201468

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #36xhigh of 2201481

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #42xhigh of 4733.8%

Leader: Gemini 4 · 68.9%

cost

AA Intelligence Index Cost per Task

177 results
  1. #25default of 177$0.05

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #32default of 177$138

Leader: GPT-6 Luna · $10.63

economics

CorpFin (Vals)

117 results
  1. #54xhigh of 11761.4%

Leader: Opus 5 · 73.2%

economics

Excel Modeling Benchmark

74 results
  1. #48xhigh of 7455.2%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #59xhigh of 7841.5%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #166default of 2104.0%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #131default of 2881121

Leader: Opus 5.5 · 1866

economics

TaxEval (Vals)

125 results
  1. #45xhigh of 12573.8%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #9default of 39979.9%
  2. #232none of 39942.7%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #201none of 52627.2%
  2. #265default of 52622.4%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #21default of 52675.3%
  2. #388none of 52610.8%

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

knowledge

GPQA Diamond

539 results
  1. #111default of 53986.6%
  2. #234none of 53976.2%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #61xhigh of 12282.6%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #80xhigh of 9432.5%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #39xhigh of 9683.7%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #60xhigh of 12284.6%

Leader: Fable 5.1 · 92.4%

long-context

AA-LCR v1.1

523 results
  1. #93default of 52379.7%
  2. #336none of 52341.7%

Leader: Kimi K3 · 88.7%

long-context

MLCR-AA

94 results
  1. #66default of 949.4%

Leader: Sonnet 5.5 · 75.0%

math

ProofBench v1.1

52 results
  1. #41xhigh of 5222.0%

Leader: Sonnet 5.5 · 100.0%

reasoning

CritPt

529 results
  1. #165default of 5294.0%
  2. #224none of 5291.1%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

LegalBench

135 results
  1. #100xhigh of 13577.1%

Leader: Fable 5 · 88.6%

reliability

Tool-call error rate

171 results
  1. #110xhigh of 1712.3%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #180xhigh of 18190.2%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #151none of 3512.0s
  2. #180default of 3512.4s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #331default of 35137.2
  2. #333none of 35136.8

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #121xhigh of 1771.9s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #162xhigh of 17629.0

Leader: Muse Spark 1.1 · 204.3