priorsBuilder

Xiaomi · released 2026-04-22

MiMo V2.5

Measured on 52 benchmarks across 3 reasoning efforts.

General frontier

41.1

#55 of 135 models · best at xhigh

includes estimates: 5% of the mix is measured for this model

Reasoning efforts in your mix

xhigh41.1#135 of 312
none36.6#173 of 312
default28.5#234 of 312

Your mix’s benchmarks

3 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #171default of 58727.2%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #185default of 526-9.8

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #111xhigh of 19635.0%
  2. #136none of 19624.0%

Leader: Opus 4.8 · 95.0%

All 52 benchmarks

show 49 more

agentic

AA-Briefcase v1.1

224 results
  1. #153default of 224754

Leader: Sonnet 5.5 · 1823

agentic

AutomationBench-AA

219 results
  1. #134default of 21918.4%

Leader: Gemini 4 · 77.5%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #49xhigh of 791.7%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #70xhigh of 789.1%

Leader: Muse Spark 1.3 · 55.3%

agentic

Tax Agent Bench

69 results
  1. #58xhigh of 697.8%

Leader: Fable 5.1 · 49.2%

agentic

τ²-Bench Telecom

393 results
  1. #61default of 39390.6%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #156default of 2068.7%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena WebDev

122 results
  1. #70xhigh of 1221438

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #56xhigh of 7714.2%

Leader: Sonnet 5.5 · 69.8%

coding

LiveCodeBench (Vals)

127 results
  1. #61xhigh of 12781.5%

Leader: Fable 5.1 · 90.5%

coding

SciCode

226 results
  1. #147default of 22643.9%

Leader: Opus 5.5 · 66.9%

coding

SWE-bench Verified (Vals)

85 results
  1. #57xhigh of 8571.0%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #39xhigh of 7460.7%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Hard-AA

387 results
  1. #52default of 38741.7%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #88default of 23863.7%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #153default of 2160.0%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #58xhigh of 10542.2%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #158default of 61025.2

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #88xhigh of 2201434

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #77xhigh of 2201451

Leader: Gemini 4 · 1553

composite

Vals Multimodal Index v1.2

33 results
  1. #23xhigh of 3352.8%

Leader: Fable 5 · 74.2%

economics

CorpFin (Vals)

117 results
  1. #70xhigh of 11759.9%

Leader: Opus 5 · 73.2%

economics

Excel Modeling Benchmark

74 results
  1. #49xhigh of 7455.1%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #66xhigh of 7836.7%

Leader: Gemini 4 · 65.4%

economics

GDPval-AA v2.1

288 results
  1. #161default of 288997

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #63xhigh of 8459.3%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #69xhigh of 12571.8%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #88default of 39967.1%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #359default of 52616.8%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #50default of 52668.1%

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

knowledge

GPQA Diamond

539 results
  1. #131default of 53984.9%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #65xhigh of 12281.6%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #83xhigh of 9431.9%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #80xhigh of 9672.2%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #71xhigh of 12282.9%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #44xhigh of 8180.0%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #96default of 26575.4%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #180default of 52373.0%

Leader: Kimi K3 · 88.7%

math

ProofBench v1.1

52 results
  1. #45xhigh of 5216.0%

Leader: Sonnet 5.5 · 100.0%

other

Arena Vision

116 results
  1. #64xhigh of 1161234

Leader: Fable 5 · 1308

reasoning

CritPt

529 results
  1. #167default of 5293.7%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

LegalBench

135 results
  1. #92xhigh of 13578.9%

Leader: Fable 5 · 88.6%

reasoning

SAGE

80 results
  1. #49xhigh of 8043.3%

Leader: Opus 4.7 · 56.1%

reliability

Tool-call error rate

172 results
  1. #148xhigh of 1725.3%

Leader: Gemini 4 · 0.0%

reliability

Uptime (7d)

182 results
  1. #180xhigh of 18292.4%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #238default of 3516.8s

Leader: Gemini 2.5 Flash Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #335default of 35135.5

Leader: Celeris-1 · 1564.1

speed

Provider latency

178 results
  1. #169xhigh of 1784.7s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #170xhigh of 17622.0

Leader: Muse Spark 1.1 · 204.5