Anthropic Releases Its Most Powerful Claude Fable Model to the Public

Anthropic has released its most advanced artificial intelligence model to the general public for the first time, but the process is being carried out with strict security measures. On Tuesday, the company launched Claude Fable 5, the public version of the Mythos model. According to Anthropic, Fable 5 delivers high performance in software engineering, intellectual labor, and visual analysis, but responses in high-risk areas such as cybersecurity, biology, and chemistry are blocked, and the system automatically switches to the Claude Opus 4.8 model. Techcrunch.com reports .

Initially introduced in April, the Mythos model was available only to limited partners due to cybersecurity concerns. Starting last week, Anthropic allowed hundreds of organizations in 15 countries to use this model. Now, this technology is accessible to everyone through Claude API and Enterprise plans. The Fable 5 model will be offered free of charge to Pro, Max, and Team subscribers until June 22, and separate credits will be required for use starting June 23.

The release of the Fable model coincides with Anthropic's preparation to enter the open market alongside giants like OpenAI and SpaceX. The company warns that artificial intelligence systems are developing so rapidly that they may soon reach the level of recursive self-improvement (RSI) without human intervention. Therefore, to prevent models at the Mythos level from falling into the wrong hands, Anthropic conducted over 1000 hours of internal and external "red-teaming" tests.

As a security measure, Anthropic has implemented a policy of retaining all traffic data for 30 days. The company emphasizes that it will not use this data to train models, but only to protect against sophisticated attacks and new types of "jailbreak" attempts. Currently, 95 percent of Fable 5 sessions are successfully completed with the model's own responses, with the system relying on Opus 4.8 assistance only in extremely complex cases.