Claude Fable 5 — Anthropic's First Mythos-Class Frontier Model, Explained

Overview

Claude Fable 5 is Anthropic's first Mythos-class frontier model, a new tier above Opus, demonstrating state-of-the-art performance across nearly all benchmarks. It features a 1-million-token context window, file-based memory for self-improvement, and a built-in safety fallback to Opus 4.8.

In this video

• Claude Fable 5 is Anthropic's first Mythos-class model, representing a new capability tier that sits above the Opus class.
• The model launched on June 9, 2026, and returned to general availability on July 1, 2026, after a brief pause due to U.S. export controls.
• Fable 5 is state-of-the-art on nearly every tested benchmark of AI capability, with its lead growing on longer and more complex tasks.
• It features a 1-million-token context window, equivalent to roughly 750,000 words or 3,000 pages, allowing for an entire mid-sized codebase in a single call.
• Fable 5 offers up to 128,000 output tokens per request, which is the largest output window Anthropic has ever offered.
• Its persistent file-based memory allows it to keep notes and improve later output, performing about three times better than Opus 4.8 in testing.
• Fable 5 posted an 80.3% score on SWE-Bench Pro, the hardest real-world software-engineering benchmark, which is the top score of any model tested.
• The model is fully multimodal, capable of processing text, image, and file inputs, and can read diagrams, charts, and tables nested inside PDFs.
• Claude Fable 5 includes built-in safeguards where prompts touching cybersecurity, biology, chemistry, or health are quietly answered by Claude Opus 4.8.
• This safety fallback triggers in under 5% of sessions on average, ensuring the full frontier model is used for everyday work.

Frequently asked questions

What is Claude Fable 5?

Claude Fable 5 is Anthropic's first Mythos-class frontier model, a new capability tier above the Opus class, demonstrating state-of-the-art performance across nearly all AI benchmarks. It became generally available on June 9, 2026.

What is the context window size of Claude Fable 5?

Claude Fable 5 features a 1-million-token context window, which translates to approximately 750,000 words or 3,000 pages, making it capable of processing very large inputs like entire codebases or extensive document corpora.

How does Claude Fable 5's persistent memory work?

Claude Fable 5's persistent file-based memory allows it to keep notes during its operation and use them to improve subsequent outputs. This feature helps it maintain focus across millions of tokens and perform about three times better than Opus 4.8.

What are Claude Fable 5's coding capabilities?

Claude Fable 5 achieved an 80.3% score on SWE-Bench Pro, the hardest real-world software-engineering benchmark, surpassing Opus 4.8. Its performance excels especially in long and complex coding tasks, useful for large-repository coding agents.

Does Claude Fable 5 have safety features?

Yes, Claude Fable 5 has built-in safety safeguards. Prompts related to high-risk topics such as cybersecurity, biology, chemistry, or health are automatically answered by Claude Opus 4.8, a more conservative model, in under 5% of sessions.