Anthropic releases Mythos version of AI model

Published 10 June 2026 · tech

Anthropic Fable 5 claimed to be better at complex and long-running tasks Anthropic has released a public version of its powerful Mythos-class artificial intelligence model

Anthropic Fable 5 claimed to be better at complex and long-running tasks Anthropic has released a public version of its powerful Mythos-class artificial intelligence model, making the technology available to a much wider audience for the first time. The new model, called Claude Fable 5, is based on the same core architecture as Claude Mythos 5, an AI system that previously drew attention from governments, cybersecurity experts and financial institutions because of its advanced capabilities. Anthropic said it decided to release the model publicly only after building extensive safeguards designed to prevent misuse in areas such as cybersecurity, biology and other high-risk fields. The company described Fable 5 as its most capable publicly available AI model to date.Announcing the new model, Anthropic said:“Releasing a model this capable comes with risks.

Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misused to cause serious damage. We’ve therefore launched the model with safeguards that mean queries on some topics will instead receive a response from our next-most-capable model, Claude Opus 4.8.To release the model both safely and quickly, we’ve tuned these safeguards conservatively—they’ll sometimes catch harmless requests, though they trigger, on average, in less than 5% of sessions”.The release is significant because Anthropic had previously limited access to Mythos-class technology through Project Glasswing, a program that provided select organizations with access to more powerful versions of the model. Anthropic said Claude Mythos 5 remains available only to trusted partners and cybersecurity experts because it has fewer restrictions.Fable 5 and Mythos 5 are being offered at $10 per million input tokens and $50 per million output tokens—less than half the price of Claude Mythos Preview.Anthropic says Fable 5 is designed for long-running and complex work that earlier AI models struggled to complete.

The company said the model performs particularly well in software engineering, research, advanced reasoning and tasks that require maintaining context over long conversations.“Fable 5 is also more token-efficient than past Claude model”, Anthropic said, adding that Fable 5 comes with a new set of classifiers: separate AI systems that detect potential misuse, including jailbreak attempts, and prevent the main model (in this case Fable 5) from responding.When Fable’s classifiers detect a request related to cybersecurity, biology and chemistry, or distillation, the response is automatically handled by Claude Opus 4.8 instead. Users will be informed whenever this occurs.The launch comes as competition in the AI industry intensifies, with Anthropic, OpenAI, Google and xAI all racing to develop more capable AI systems while addressing growing concerns about safety and misuse.Anthropic further stated that it will require 30-day retention for all traffic on Mythos-class models, on both first- and third-party surfaces.