On June 9, 2026, Anthropic announced Claude Fable 5 and Claude Mythos 5. The June 9, 2026 date is stated on the official Anthropic announcement. Fable 5 is a Mythos-class model built for general public use and described as Anthropic’s most capable widely released model, strong in software engineering, knowledge work, vision, and scientific research. Mythos 5 is the same underlying model with certain safety restrictions removed, restricted to cyberdefenders and infrastructure providers through Project Glasswing, with planned expansion to biomedical researchers via a trusted access program (both descriptions trace to the primary).
Anthropic reports that Fable 5 posts the highest scores on Cognition’s FrontierCode coding evaluation and is the first model to break 90% on a complex analytics benchmark, a 10-point jump (claims trace to the primary). Pricing is $10 per million input tokens and $50 per million output tokens, which the post notes is less than half the cost of Claude Mythos Preview (numbers trace to the primary). Availability is immediate via the Claude API and subscription plans, included on Pro, Max, Team, and seat-based Enterprise plans from launch through June 22, with credit-based usage after June 23 (dates trace to the primary). When classifiers detect requests related to cybersecurity, biology, chemistry, or distillation, the system automatically falls back to Claude Opus 4.8, which Anthropic says affects less than 5% of sessions (claim traces to the primary).
For business and technical leaders, Fable 5 matters both as a capability jump for long-horizon agentic work and as a case study in safety-gated deployment. Shipping a powerful general model alongside an access-controlled variant, with automatic fallback on sensitive domains, shows how frontier labs are trying to release capability while containing dual-use risk.