AIFIRST Claude Fable 5.1 and Claude Mythos 5.1
The brief
Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, its most capable models for coding and long-running agentic tasks.
Key points
- Fable 5.1 scores 52.6% on Terminal-Bench-Science and 73.4% on CursorBench, surpassing all prior Claude models.
- Pricing drops roughly 25% for typical workloads and 45% for agentic use, with cache read costs cut 75%.
- Mythos 5.1 is the same model with restricted access and additional enterprise safety controls (API ID: claude-fable-5-1).
Sources
- HNanthropic.com