Mistral Small 4

Composite indexes

• Artificial Analysis Intelligence Index#94 of 142
  11independentIntelligence Index v4.3; reasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗
  9independentIntelligence Index v4.3; non-reasoning modeArtificial Analysis: Mistral Small 4 (Non-reasoning) ↗

  Leader: Claude Opus 5.5 at 58

Coding

• SciCode#73 of 94
  38.8%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Opus 5.5 at 66.9%

• Terminal-Bench 4.0#76 of 88
  0%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Mythos 5.1 at 60.9%

• Terminal-Bench 2.x#82 of 103
  21%independentTerminal-Bench 2.1; reasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Fable 5.1 at 91.4%

Agents and tool use

• τ²-Bench#50 of 88
  41.2%independenttelecom; reasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: GLM-5.2 at 99.1%

• GDPval-AA#84 of 103
  349independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Opus 5.5 at 1846

Reasoning and knowledge

• GPQA Diamond#79 of 142
  76.9%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: GPT-6 Astra at 96.1%

• Humanity’s Last Exam#106 of 147
  9.9%independentno tools; reasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Opus 5.5 at 61.4%

• CritPt#80 of 139
  0.3%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: GPT-5.6 Sol at 32.3%

Vision and long context

• AA Long Context Reasoning#92 of 139
  49.7%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Kimi K3 at 88.7%

• MMMU-Pro#60 of 72
  56.8%independentreasoningArtificial Analysis: Mistral Small 4 (Reasoning) ↗

  Leader: Claude Opus 5.5 at 87.7%

Other published results