Grok Build 0.1

Composite indexes

• Artificial Analysis Intelligence Index#45 of 142
  27independentAA Intelligence Index v4.3.2, defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Claude Opus 5.5 at 58

Coding

• Terminal-Bench 2.x#59 of 103
  52.1%independentTerminal-Bench 2.1 (AA harness), defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Claude Fable 5.1 at 91.4%

Agents and tool use

• GDPval-AA#48 of 103
  1052independentGDPval-AA, defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Claude Opus 5.5 at 1846

Reasoning and knowledge

• Humanity’s Last Exam#41 of 147
  38.3%independentno tools (AA), defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Claude Opus 5.5 at 61.4%

• GPQA Diamond#43 of 142
  89.5%independentAA, defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: GPT-6 Astra at 96.1%

• CritPt#43 of 139
  9.1%independentdefaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: GPT-5.6 Sol at 32.3%

Vision and long context

• AA Long Context Reasoning#59 of 139
  74.7%independentdefaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Kimi K3 at 88.7%

• MMMU-Pro#30 of 72
  76.5%independentAA, defaultArtificial Analysis: Grok Build 0.1 0616 ↗

  Leader: Claude Opus 5.5 at 87.7%

Human preference

• LMArena WebDev#54 of 54
  1165independentgrok-code-fast-1 (now alias of grok-build-0.1; may reflect the older model), LMArena Code arena (formerly WebDev), vote cutoff 2026-09-23LMArena Code (WebDev) leaderboard ↗

  Leader: Claude Opus 5.5 at 1818

Other published results