Claude Fable 5.1

Composite indexes

• Artificial Analysis Intelligence Index#4 of 151
  53independentv4.3.2, max effortArtificial Analysis: Claude Fable 5.1 ↗
  51independentv4.3.2, high effort (API default)Artificial Analysis: Claude Fable 5.1 (high) ↗

  Leader: Claude Opus 5.5 at 58

Coding

• SciCode#2 of 100
  63.1%independentAA run, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Claude Opus 5.5 at 66.9%

• Terminal-Bench 4.0#5 of 96
  57.9%independentmax effort, Claude Code harness, ±3.8Terminal-Bench 4.0 official leaderboard ↗
  55.8%vendormax effort, Claude Code harness (Mythos 5.1 same model: 60.9%)Anthropic: Claude Fable 5.1 and Mythos 5.1 ↗
  52%independentAA run (Terminal-Bench 4.0), max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Claude Sonnet 5.5 at 70.6%

• Terminal-Bench 2.x#1 of 105
  91.4%independentAA run (Terminal-Bench 2.1), max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Claude Fable 5.1 at 91.4%

• SWE-bench Pro#3 of 33
  81.2%vendorAnthropic-run, max effort, avg of 5 trials (set not specified)Claude Fable 5.1 / Mythos 5.1 System Card, Table 8.1.A ↗

  Leader: Claude Opus 5.5 at 89.9%

Agents and tool use

• GDPval-AA#4 of 109
  1735independentGDPval-AA v2.1, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Muse Wonder Full at 1857

• OSWorld-Verified#9 of 21
  77.9%vendorOSWorld 2.0 partial score (strict 41.7%), max effort; Opus 5.5 system card later reports 80.7/42.8 under a revised harnessAnthropic: Claude Fable 5.1 and Mythos 5.1 ↗

  Leader: Kimi K3 at 84.8%

Reasoning and knowledge

• GPQA Diamond#10 of 144
  93.7%independentAA run, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: GPT-6 Astra at 96.1%

• Humanity’s Last Exam#3 of 153
  65%vendorwith tools, max effort (Opus 5.5 post later lists 65.6%)Anthropic: Claude Fable 5.1 and Mythos 5.1 ↗
  60.9%vendorno tools, max effortAnthropic: Claude Fable 5.1 and Mythos 5.1 ↗
  59.1%independentAA run, no tools, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Claude Sonnet 5.5 at 64.5%

• CritPt#7 of 144
  29.7%independentAA run, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Muse Wonder Full at 33%

• ARC-AGI-2#5 of 25
  90%independentmax effort; ARC Prize verified, cost $4.49/taskARC Prize leaderboard data (semi-private) ↗

  Leader: GPT-6 Astra at 95%

Vision and long context

• AA Long Context Reasoning#4 of 144
  85.3%independentAA-LCR v1.1, max effortArtificial Analysis: Claude Fable 5.1 ↗

  Leader: Kimi K3 at 88.7%

Human preference

• LMArena Text#4 of 57
  1499independentclaude-fable-5.1-max; 5,783 votesLMArena text leaderboard ↗

  Leader: Gemini 4 Argon at 1525

• LMArena WebDev#4 of 58
  1755independentclaude-fable-5.1-max; 4,916 votesLMArena Code Arena | WebDev leaderboard ↗

  Leader: Claude Opus 5.5 at 1818

Other published results