Type: AI model
Concepts: LLM Evals Mentioned alongside: AISI Anthropic GPT-6 Astra Recent briefs: 2026-10-23
Recent Updates
- 2026-07-24: Claude Opus 5 launches with stronger coding and knowledge-work benchmark results (Hacker News) · anthropic.com — Agents Code Agents LLM Evals Claude Opus 5 · Claude Fable 5 · Opus 4.8
- 2026-07-30: Investigating three real-world incidents in our cybersecurity evaluations (Hacker News) · anthropic.com — LLM Evals Anthropic · OpenAI · Hugging Face · Irregular · Opus 4.7
- 2026-08-14: GLM-5.3 Released: Open-Source Coding Model Approaches Fable 5 Level, Becomes Strongest Open-Source Security Model (量子位) · qbitai.com — Code Agents Open Source LLMs LLM Evals Zhipu · Anthropic · OpenAI · Tsinghua University Nankai University Karpathy · GLM-5.3 · GLM-5.2 · Fable 5 · Opus 4.8 · GPT-5.6 Sol
- 2026-09-02: Fable 5.1 and Mythos 5.1 Released with 45% Cost Reduction and Anti-Distillation Security (breakingnewsofficial) · qbitai.com — Agents Anthropic · OpenAI · NASA Adaptyv Bio ESA Flix Rieseberg Ilya · Fable 5.1 Mythos 5.1 · Fable 5 · GPT-5.6 Sol · Opus
- 2026-10-23: AISI and Anthropic Report Unsanctioned Supply-Chain Attacks by Models During Cybersecurity Evaluations (breakingnewsofficial) · aisi.gov.uk — LLM Evals Anthropic · AISI · GPT-6 Astra
FAQ
What is Mythos 5?
Mythos 5 is the shorthand for Anthropic’s Claude Mythos 5, the partner-only, less-restricted sibling of Claude Fable 5 in the fifth Claude generation. It shares Fable 5’s base model but is reserved for select partners pending additional safeguards. GROUNDING tracks Mythos 5’s restricted availability and relationship to Fable 5.
What does this page track?
Dated radar mentions, source links, related concepts, and builder-relevant context for Mythos 5, collected automatically by GROUNDING.
When was Mythos 5 last mentioned?
Mythos 5 was most recently mentioned in a radar update dated 2026-10-23.
Category: Text / Language Models