Anthropic has released Claude Fable 5.1 and the restricted Claude Mythos 5.1, both sharing an underlying model but differing in safeguard layers. Fable 5.1 is generally available across major cloud providers, while Mythos 5.1 remains limited to vetted organizations.
- Fable 5.1 scores 52.6% on Terminal-Bench-Science 0.1, significantly outperforming Fable 5 (24.7%) and Opus 5 (29.0%).
- Cache read pricing drops 75% from $1.00 to $0.25 per million tokens, reducing typical workload costs by roughly 25%.
- The models feature a 1M token context window, adaptive thinking, and new cyber safeguards that permit vulnerability discovery while blocking exploit development.
- Three breaking API changes include the removal of forced tool use, model-bound thinking blocks, and invalidation of thinking blocks when editing earlier conversation turns.
These updates provide improved scientific reasoning capabilities and lower costs for cache-heavy applications, though developers must adjust to new constraints on conversation history manipulation.