Anthropic Releases Fable 5.1 with Massive Gains
AGI Hunt · wechat · 2026-09-02
Anthropic released Fable 5.1 and Mythos 5.1.
Performance
- Science: 52.6% on Terminal-Bench-Science 0.1 (up from 24.7%).
- Coding: 56% on Terminal-Bench 4.0 (Mythos 5.1: 61%); 73.4% on CursorBench 3.2.0.
- Reasoning: 60.9% on Humanity's Last Exam (no tools).
- Computer Use: 41.7% strict score on OSWorld 2.0.
- Automation: 31.4% on AutomationBench.
PreservedThinking
- New anti-distillation protection validates thinking block context integrity. Requests fail or drop thinking blocks if context is tampered.
- Restriction applies only to new API accounts created after Aug 31, 2026.
Related event: Anthropic Launches Claude Fable 5.1 with 75% Cache Price Cut(72 posts)→
More from Models
- OpenAI previews Astra, a cybersecurity model reaching 'Critical' threshold — mikegiannulis · 2026-09-02
- Dev: Anthropic 5.1 is a massive leap over previous model — doodlestein · 2026-09-02
- Fable 5.1 test fails to meet expectations — Angaisb_ · 2026-09-02
- Claude Fable 5.1 generates a full Mario Kart game with a single prompt — ezshine · 2026-09-02
- Gemini 3.7 Flash Demonstrates Agentic Video Understanding — otarU · 2026-09-02
- OpenAI's Astra achieves 100% success rate on ExploitBench vulnerability tests — JiaweiLiu_ · 2026-09-02