MiMo-V2.5

Composite indexes

• Artificial Analysis Intelligence Index#53 of 142
  25independentv4.3.2 (AA estimate; independent evaluation forthcoming)Artificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Opus 5.5 at 58

Coding

• Terminal-Bench 4.0#71 of 88
  0%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Mythos 5.1 at 60.9%

• Terminal-Bench 2.x#45 of 103
  63.7%independentAA, Terminal-Bench 2.1Artificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Fable 5.1 at 91.4%

• SciCode#60 of 94
  43.9%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Opus 5.5 at 66.9%

Agents and tool use

• τ²-Bench#18 of 88
  90.6%independentAA, telecomArtificial Analysis: MiMo-V2.5 ↗

  Leader: GLM-5.2 at 99.1%

• GDPval-AA#55 of 103
  986independentGDPval-AA v2.1Artificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Opus 5.5 at 1846

Reasoning and knowledge

• Humanity’s Last Exam#61 of 147
  27.2%independentAA, no toolsArtificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Opus 5.5 at 61.4%

• GPQA Diamond#59 of 142
  84.9%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: GPT-6 Astra at 96.1%

• CritPt#55 of 139
  3.7%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: GPT-5.6 Sol at 32.3%

Vision and long context

• AA Long Context Reasoning#66 of 139
  73%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: Kimi K3 at 88.7%

• MMMU-Pro#34 of 72
  75.4%independentAAArtificial Analysis: MiMo-V2.5 ↗

  Leader: Claude Opus 5.5 at 87.7%

Human preference

• LMArena Text#42 of 54
  1434independentarena entry "mimo-v2.5", rank 96, 44466 votesLMArena Text leaderboard (default view, fetched 2026-09-25) ↗

  Leader: Claude Fable 5 at 1506

• LMArena WebDev#44 of 54
  1437independentarena entry "mimo-v2.5", rank 68, 14099 votesLMArena WebDev leaderboard (fetched 2026-09-25) ↗

  Leader: Claude Opus 5.5 at 1818

Other published results