Anthropic Launches Claude Fable 5, the First Public Mythos-Tier AI Model

Anthropic has officially launched Claude Fable 5, its first publicly accessible Mythos-tier AI model. The company states that Fable 5 has passed safety evaluations and is suitable for general-purpose applications. According to Anthropic, this model surpasses all previous public releases, demonstrating "exceptional performance" in areas such as software engineering, knowledge work, visual recognition, and scientific research. The launch of Fable 5 marks a major step forward in Anthropic's push toward general AI capabilities.

Notably, most queries related to cybersecurity, chemistry, and biology will be handled by Opus 4.8 rather than Fable 5—a design intended to keep responses in sensitive domains more controlled. Meanwhile, Anthropic has also deployed Mythos 5 via Project Glasswing, serving as an upgrade to the Claude Mythos Preview. Mythos 5 is described as having the world's strongest cybersecurity capabilities and is currently available only to users who already have access to the Mythos Preview, including Apple. Anthropic plans to gradually expand Mythos 5's availability through a broader trusted access program.

On the same day, Anthropic also announced the release of its latest model, Claude Opus 4.8. The tiered strategy of Fable 5 and Mythos 5 reveals Anthropic's intent to balance high-performance general-purpose models with high-safety sensitive models.

Analysis: The release of Fable 5 further enriches Anthropic’s product lineup. Its "safe and usable" positioning is likely to attract more enterprise-level users. Meanwhile, Mythos 5's specialization in cybersecurity highlights the deep application potential of AI in vertical security scenarios. Going forward, as the trusted access program expands, the Mythos series could become the benchmark for industry security AI.