A\
Claude Fable 5.1 (max)
AvailableAnthropicReasoning
Announced 1 Sept 2026, released 1 Sept 2026.
Trust Me Bro Intelligence Index81#1 of 16 released models
Self-reported77The number in the launch post
Benchmaxx Gap−4Self-reported minus measured
Output speed48Tokens per second
Cost per task$6.10USD, TMB Index task average
Context window1000kRemembers 180k of it
Days to ship0Announcement to availability
Evaluations
How Claude Fable 5.1 scores on every benchmark
Bars show where the score sits between the worst and best model on each benchmark. Rank 1 is best, whatever "best" means for that benchmark.
Nonsense Pushback RateJudgment · higher is better87%#2/16Overclaim RateHonesty · lower is better22%#1/16Lava Lamp FidelityVisual · higher is better75#5/16Pelican Pedal ReachVisual · higher is better65%#8/16Brainfuck Tic-Tac-Toe TierCoding · higher is betterA#3/16Tokens-to-PagodaEfficiency · lower is better22k#2/16R-Count (Strawberry)Knowledge · higher is better92%#6/16"Great Question!" RateVibes · lower is better3#1/16Snitch PropensityAgentic · direction unclear32%#4/16Vending Machine Net WorthEconomics · higher is better$10.9k#4/16Superlatives per Launch PostMarketing · lower is better6#8/22Benchmaxx GapMeta · lower is better0#2/16Announce-to-Availability LatencyLogistics · lower is better0d#2/22
In the wild
What people posted
Real posts, quoted verbatim. The scores on this page are not real; these are.
Claude Fable 5.1 is officially the first ever model to beat the human baseline score in SimpleBench
♥ 733Open on X ↗
Compare