Anthropic announced the launch of Claude Fable 5 and Claude Mythos 5, introducing the first publicly available models from its new Mythos class of AI systems. The release marks a significant milestone for the company, making its most powerful technology available to a broader set of users while maintaining safeguards designed to limit misuse.
Claude Fable 5 is the public-facing version of the Mythos architecture and is designed for software engineering, knowledge work, and visual reasoning tasks. Anthropic said the model delivers state-of-the-art performance and becomes increasingly effective as tasks grow longer and more complex. The company noted that new safety mechanisms allow the model to block responses in certain high-risk domains and automatically fall back to Claude Opus 4.8 when necessary.
Alongside Fable 5, Anthropic introduced Claude Mythos 5, which shares the same underlying architecture but offers fewer restrictions. Access to Mythos 5 is currently limited to approved organizations participating in Project Glasswing, a program that includes cybersecurity and infrastructure partners. Anthropic plans to expand access gradually through a trust-based model.
Both models feature a one million-token context window and support up to 128,000 output tokens. Anthropic said adaptive reasoning is always enabled, allowing users to adjust the depth of the model’s reasoning without exposing its raw chain of thought.
Pricing for Claude Fable 5 and Claude Mythos 5 is set at $10 per million input tokens and $50 per million output tokens. The models are available through the Claude API, Amazon Bedrock, Vertex AI, and Microsoft Foundry.
Anthropic emphasized that the launch was made possible by the development of safeguards designed to address concerns surrounding cybersecurity and biological research. Earlier versions of the Mythos family were not released publicly because of concerns about their advanced capabilities in those areas.
The company said Fable 5 will initially be included in Pro, Max, Team, and enterprise plans before transitioning to a usage-based model as Anthropic manages capacity constraints.
Anthropic, founded in 2021 by former OpenAI researchers, focuses on developing advanced AI systems with an emphasis on safety and responsible deployment. The introduction of the Mythos family establishes a new class of models above the company’s Opus line and represents its most capable AI technology to date.
KEY QUOTES:
“Marketing teams are under pressure to deliver better outcomes with more complexity, more channels and more data than ever before. Minerva gives marketers the context and infrastructure to understand their customers more deeply and act on those insights faster. Our goal is to give repetitive, operational work to AI so our customers can spend more time on the work that requires real human judgment.”
Jackson Engles, Co-Founder And CEO, Minerva
“AI agents are context-hungry, and whoever structures the right context for a domain wins that domain. Minerva is building the context layer for marketing. Most brands have valuable first-party data, but it’s fragmented and hard to use. Minerva turns that data into something AI can actually reason over, so then marketers can act on it.”
Phin Barnes, Co-Founder And Managing Partner, The General Partnership
