Grok 4.6

Composite indexes

• Artificial Analysis Intelligence Index#13 of 142
  44independentAA Intelligence Index v4.3.2, highArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Claude Opus 5.5 at 58

Coding

• Terminal-Bench 4.0#21 of 88
  21.2%independentTerminal-Bench 4.0 (AA harness), highArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Claude Mythos 5.1 at 60.9%

• SciCode#19 of 94
  56.5%independenthighArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Claude Opus 5.5 at 66.9%

Agents and tool use

• GDPval-AA#11 of 103
  1606independentGDPval-AA, highArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Claude Opus 5.5 at 1846

Reasoning and knowledge

• Humanity’s Last Exam#24 of 147
  42.9%independentno tools (AA), highArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Claude Opus 5.5 at 61.4%

• GPQA Diamond#6 of 142
  94.9%independentAA, highArtificial Analysis: Grok 4.6 (high) ↗

  Leader: GPT-6 Astra at 96.1%

• CritPt#25 of 139
  17.1%independenthighArtificial Analysis: Grok 4.6 (high) ↗

  Leader: GPT-5.6 Sol at 32.3%

• ARC-AGI-2#13 of 25
  67.1%independentsemi-private, XHighARC Prize leaderboard ↗

  Leader: GPT-6 Astra at 95%

• ARC-AGI-3#5 of 9
  2.1%independentsemi-private, XHighARC Prize leaderboard ↗

  Leader: GPT-6 Astra at 62.7%

Vision and long context

• AA Long Context Reasoning#26 of 139
  80.3%independenthighArtificial Analysis: Grok 4.6 (high) ↗

  Leader: Kimi K3 at 88.7%

Human preference

• LMArena Text#28 of 54
  1456independentgrok-4.6-high, style control default, vote cutoff 2026-09-13LMArena Text leaderboard ↗

  Leader: Claude Fable 5 at 1506

• LMArena WebDev#13 of 54
  1624independentgrok-4.6-high, LMArena Code arena (formerly WebDev), vote cutoff 2026-09-23LMArena Code (WebDev) leaderboard ↗

  Leader: Claude Opus 5.5 at 1818

Other published results