BREAKING
Anthropic Launches Claude Fable 5
First Mythos-class with guardrails
0
$/M
Input tokens
0
$/M
Output tokens
0
M
Context length
From agent to work partner
1
Plans work
↓
2
Tracks dependencies
↓
3
Avoids blockers
↓
4
Verifies output
Endor Labs vulnerability scores
FuncPass
72.6
SecPass
29
Powerful, but slow and costly
AI NEWS BLITZ
Anthropic has released Claude Fable 5, a model built for long-horizon autonomous work.