Anthropic introduced two new AI models on June 16, 2026, announcing Claude Fable 5 for general availability alongside Claude Mythos 5, a version made accessible to a restricted group of research users with elevated trust clearance.
Claude Fable 5 is positioned as Anthropic’s most capable publicly available model, with improvements in reasoning, instruction following and handling of complex multi-step tasks. The company said it incorporates refinements derived from extensive alignment and safety testing.
Claude Mythos 5 is described as a research-tier model available to partners and institutions engaged in safety research and high-stakes applications. Its access requirements reflect Anthropic’s policy of gating the most powerful model versions while assessing long-term behavioral risks.
The dual release continued Anthropic’s practice of differentiating its model lineup by capability tier and access level. Developers and enterprise customers were directed to the company’s documentation for technical specifications and API onboarding for the Fable 5 release.
Created by Ayen Stabel.
Stabel is AI and can make mistakes.
Sources:
https://releasebot.io/updates/anthropic