MiMo-V2.5-Pro

Composite indexes

• Artificial Analysis Intelligence Index#50 of 142
  26independentv4.3.2, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Opus 5.5 at 58

Coding

• Terminal-Bench 4.0#70 of 88
  0%independentAA, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Mythos 5.1 at 60.9%

• Terminal-Bench 2.x#42 of 103
  65.2%independentAA, Terminal-Bench 2.1, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Fable 5.1 at 91.4%

• SciCode#38 of 94
  50.6%independentAA, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Opus 5.5 at 66.9%

Agents and tool use

• τ²-Bench#13 of 88
  94.2%independentAA, telecom, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: GLM-5.2 at 99.1%

• GDPval-AA#46 of 103
  1107independentGDPval-AA v2.1, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Opus 5.5 at 1846

Reasoning and knowledge

• Humanity’s Last Exam#45 of 147
  35.7%independentAA, no tools, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Claude Opus 5.5 at 61.4%

• GPQA Diamond#52 of 142
  86.6%independentAA, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: GPT-6 Astra at 96.1%

• CritPt#53 of 139
  4%independentAA, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: GPT-5.6 Sol at 32.3%

Vision and long context

• AA Long Context Reasoning#36 of 139
  79.7%independentAA, reasoningArtificial Analysis: MiMo-V2.5-Pro ↗

  Leader: Kimi K3 at 88.7%

Human preference

• LMArena Text#22 of 54
  1467independentarena entry "mimo-v2.5-pro", rank 44, 60919 votesLMArena Text leaderboard (default view, fetched 2026-09-25) ↗

  Leader: Claude Fable 5 at 1506

• LMArena WebDev#40 of 54
  1476independentarena entry "mimo-v2.5-pro", rank 55, 17980 votesLMArena WebDev leaderboard (fetched 2026-09-25) ↗

  Leader: Claude Opus 5.5 at 1818

Other published results