Anthropic Unveils Claude Fable 5 for Creative AI Tasks
Meet Anthropic's Claude Fable 5, the first public Mythos-family AI model, combining stronger reasoning with strict safety guardrails.
A fresh chapter in creative AI begins today. Anthropic has introduced Claude Fable 5, the first publicly available model in its Mythos family, aimed at creative work alongside software engineering, knowledge tasks and image understanding.
Announced on 10 June, the release is framed as a step up in reasoning over longer, more complex briefs, with pricing set at roughly twice the cost of Claude Opus 4.8.
Designed to help creators and coders deliver
Fable 5 targets real-world workflows such as drafting stories, generating design concepts, producing product documentation and assisting with code. Anthropic says the model handles multi-stage problems and visual inputs more reliably than earlier Claude versions. According to cited benchmarks, it currently leads publicly available AI systems and posts an overall score about 5% higher than Claude Opus 4.8.
Safety boundaries with a clear fallback path
The public release applies strict guardrails in high-risk areas, including cybersecurity, biology, chemistry and model distillation, which is the process of compressing one model into another.
If a user requests restricted actions, Fable 5 blocks the task and falls back to Claude Opus 4.8. Anthropic will also offer a related variant with fewer restrictions, called Mythos 5, to vetted groups through Project Glasswing.
How the guardrails were tested
Before launch, Anthropic ran an external bug bounty programme and invited red team testers to probe for weaknesses. After more than 1,000 hours of attempts, the company reported no universal jailbreak that could consistently bypass the protections, pointing to stronger resilience than previous setups. In this context, red teaming means specialists trying to make the model behave outside its intended limits.
Rollout, access and who benefits first
Access to the cyber-capable Mythos track is expanding through Project Glasswing to 150 organisations across 15+ countries, including India, as well as select partners such as members of the Five Eyes alliance and NATO.
Anthropic says it plans to widen eligibility over time. For creative teams and product engineers, this staged rollout clarifies where advanced capabilities will appear first.
What it means for everyday creative work
For studios, marketers and developers, Fable 5’s blend of stronger reasoning and improved visual understanding could speed up ideation, drafting and code support while keeping sensitive domains off limits.
The fallback to Opus 4.8 may limit some research uses, but it also sets a clear boundary. As Glasswing expands, guidance on choosing between Fable 5 for general creativity and Mythos 5 for specialised, approved contexts should become more concrete.


