Olmo 3.1 32B Think

Composite indexes

• Artificial Analysis Intelligence Index#122 of 142
  7independentIntelligence Index v4.3; reasoning; estimatedArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: Claude Opus 5.5 at 58

Coding

• LiveCodeBench#25 of 56
  69.5%independentreasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: Fugu at 92.9%

Agents and tool use

• τ²-Bench#85 of 88
  0%independenttelecom; reasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: GLM-5.2 at 99.1%

Reasoning and knowledge

• GPQA Diamond#111 of 142
  59.1%independentreasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: GPT-6 Astra at 96.1%

• Humanity’s Last Exam#114 of 147
  6.3%independentno tools; reasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: Claude Opus 5.5 at 61.4%

• CritPt#124 of 139
  0%independentreasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: GPT-5.6 Sol at 32.3%

Vision and long context

• AA Long Context Reasoning#131 of 139
  0%independentreasoningArtificial Analysis: Olmo 3.1 32B Think ↗

  Leader: Kimi K3 at 88.7%

Other published results