Mistral Large 3

Composite indexes

• Artificial Analysis Intelligence Index#111 of 151
  9independentIntelligence Index v4.3; non-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Opus 5.5 at 58

Coding

• SciCode#85 of 100
  36.6%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Opus 5.5 at 66.9%

• Terminal-Bench 4.0#85 of 96
  0%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Sonnet 5.5 at 70.6%

• Terminal-Bench 2.x#93 of 105
  12%independentTerminal-Bench 2.1; non-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Fable 5.1 at 91.4%

• LiveCodeBench#36 of 57
  46.5%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Fugu at 92.9%

Agents and tool use

• τ²-Bench#63 of 89
  24.6%independenttelecom; non-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: GLM-5.2 at 99.1%

• GDPval-AA#88 of 109
  406independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Muse Wonder Full at 1857

Reasoning and knowledge

• GPQA Diamond#103 of 144
  68%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: GPT-6 Astra at 96.1%

• Humanity’s Last Exam#142 of 153
  4.2%independentno tools; non-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Sonnet 5.5 at 64.5%

• CritPt#110 of 144
  0%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Muse Wonder Full at 33%

Vision and long context

• AA Long Context Reasoning#106 of 144
  36%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Kimi K3 at 88.7%

• MMMU-Pro#61 of 72
  55.7%independentnon-reasoningArtificial Analysis: Mistral Large 3 ↗

  Leader: Claude Opus 5.5 at 87.7%

Other published results