BREAKING
Anthropic Launches Claude Fable 5
0M
context tokens
0k
max output
SWE-Bench Pro-equivalent score
Fable 580.3
Opus 4.869.2
GPT-5.558.6
Fable 5 vs Mythos 5
Claude Fable 5Public
Strengthened guardrails
Input $10 / Output $50 per 1M
Claude Mythos 5Partners
Project Glasswing only
Some safety measures removed
Safety vs utility debate
Balancing safety and utility
AI NEWS BLITZ
Anthropic has unveiled Claude Fable 5, its most capable public model yet.