Anthropic launches Claude 5.1: Fable and Mythos tiers with 128k output and adaptive thinking
btibor91 · x · 2026-09-02
Anthropic introduced Claude Fable 5.1 and Mythos 5.1, both featuring a 1M-token context window, up to 128k output tokens, and always-on adaptive thinking. Fable 5.1 is generally available for demanding reasoning and long-running agentic workflows, outperforming Opus 5, while Mythos 5.1 is restricted to cybersecurity and life sciences via trusted access. Updates include lower cache-read pricing, changes to API thinking blocks, new enterprise data controls, and content provenance features. Mythos 5.1 demonstrated significant capabilities in protein binder design.
Related event: Anthropic Releases Claude Fable 5.1 with 75% Cache Price Cut(90 posts)→
More from Models
- GLM Flash outperforms Opus 4.6 Max in coding tasks — 9r4n4y · 2026-09-02
- MiniMax-H3 Shows Strong Controllability, Needs Only 0.2% Fine-tuning — Hailuo_AI · 2026-09-02
- GLM 5.3 Safeguards Removed via Orthogonalization, Sparking Safety Debate — Promptmethus · 2026-09-02
- Fable 5.1 Usage Test: Only 29% Consumed After Heavy Use — evielync · 2026-09-02
- Hermes group chat bot speedup confirmed by developer — Teknium · 2026-09-02
- Moonshot Deprecates moonshot-v1, Recommends Kimi K3 — emmanuelvivier · 2026-09-02