Anthropic has officially introduced Claude Fable 5, a next-generation artificial intelligence model that the company describes as its most capable publicly available Claude model to date. The release marks the first public deployment of Anthropic’s new Mythos-class technology, designed to handle complex reasoning, software engineering, research, and long-duration analytical tasks.
Claude Fable 5 is based on the same underlying architecture as Anthropic’s more advanced Mythos 5 model. However, Fable 5 includes additional safety mechanisms intended to prevent misuse in sensitive domains such as cybersecurity, biology, and chemistry. When users attempt to access restricted capabilities, the system can redirect requests to the older Claude Opus 4.8 model.
According to Anthropic, the new model demonstrates major improvements in software development, scientific research, business analysis, and other multi-step workflows. The company says Fable 5 can autonomously perform tasks that previously required hours or even days of human effort.
One of the model’s key upgrades is its ability to process significantly larger amounts of information. Claude Fable 5 supports a context window of up to one million tokens, enabling users to work with extensive codebases, large documents, and complex research projects more efficiently.
Alongside Fable 5, Anthropic also announced Mythos 5, a version with fewer restrictions that shares the same core capabilities. Access to Mythos 5 is currently limited to selected organizations and trusted partners participating in Anthropic’s advanced research and security programs.
Claude Fable 5 is available through Anthropic’s API and enterprise offerings. Pricing has been set at $10 per million input tokens and $50 per million output tokens, reflecting the model’s significantly enhanced capabilities compared to previous Claude generations.
The launch comes at a time of intensifying competition across the AI industry. With Fable 5, Anthropic aims to strengthen its position against rivals including OpenAI, Google, and xAI, while continuing to emphasize safety and responsible deployment of frontier AI systems.
