AnthropicProvisional

Claude Fable 5

Claude 5 / Fable · claude-fable-5

Anthropic high-capability Claude model for the hardest agentic and coding workloads.

Objective score

43.3

Confidence

82%

Price evidence

Pricing evidence pending

Support this model Compare trend Official source

Other primary tiers in this family

Scenario evidence

Six task-specific rankings

Feature development#18Obj. 43.30 · 0%Debug and repair#13Obj. 55.80 · 0%Refactoring#14Obj. 55.20 · 0%Code review#11Obj. 59.90 · 0%Large-repo agent work#11Obj. 47.40 · 0%Frontend / UI#15Obj. 55.20 · 0%

Evidence state

Signal coverage

Benchmarks: 2
Rank points: 7
Usage: 2
Agent: 1
Sources: 5

Objective evidence

Benchmarks and public signals

Swebench: —
Livecodebench: —
Aider: —
Artificial Analysis: 81.6
Openrouter Usage: 0.7
Kilo Usage: —

Persistent history

Recent ranking points

Sep 15#20 · 41

Sep 15#20 · 41

Sep 15#20 · 41

Sep 14#20 · 41

Sep 14#20 · 41

Sep 14#20 · 41

Usage signals

Overall0.8%

OpenRouter0.7%

Agent and tool signals

Claude Code · 62.2

Claude Code - Fable 5.1 (max) (with fallback)

Source trail

Anthropic model overview names Claude Fable 5 as the highest capability current Claude model.Artificial Analysis Coding Agents · ok WebDev Arena · ok Claude Platform models · ok Kilo leaderboard · ok OpenRouter rankings · ok