Anthropic has officially launched Claude Fable 5, positioning it as the company’s most capable generally available AI model to date. Alongside Fable 5, the company introduced Claude Mythos 5, a specialized version operating with fewer safeguards in specific domains. The announcement on June 9, 2026, marks a significant leap in AI capability, with both models excelling across benchmarks in software engineering, knowledge work, vision, and scientific research.
What Fable 5 Brings to the Table
Fable 5 has demonstrated state-of-the-art performance across numerous AI capability benchmarks, particularly in tasks requiring extended and complex reasoning. Its proficiency spans diverse fields: intricate software engineering challenges, advanced scientific inquiry, and sophisticated knowledge work.
The pricing structure makes Fable 5 accessible at scale. Input tokens cost $10 per million, while output tokens run $50 per million, making it less than half the cost of its predecessor, Claude Mythos Preview.
Claude Mythos 5: Power Without the Guardrails
Claude Mythos 5 shares the same underlying model as Fable 5 but operates with fewer safeguards in specific domains. The version is lauded for unparalleled cybersecurity capabilities and has been deployed through Project Glasswing in collaboration with the US government, serving cyberdefenders and infrastructure providers.
This dual approach allows Anthropic to deliver cutting-edge capability where it matters most while maintaining broader safety standards for general-purpose access.
Safety Built In, Not Bolted On
Releasing such powerful models requires stringent safety protocols. Anthropic has implemented conservative safeguards for Fable 5 that occasionally redirect queries on sensitive topics to Claude Opus 4.8, the company’s next-most-capable model. These safeguards trigger in less than 5% of sessions on average and are specifically designed to prevent misuse in areas like cybersecurity and biology.
The company acknowledges that these measures may sometimes flag harmless requests but is actively working to refine them and reduce false positives in the coming months. This iterative approach reflects a commitment to balancing capability with responsibility.
Real-World Results Already Coming In
Engineering at scale: Stripe reported that Fable 5 compressed months of engineering work into days. One instance saw a codebase-wide migration completed in a single day, a task that would typically take a team over two months. On Cognition’s FrontierCode evaluation, Fable 5 achieved the highest scores among frontier models for difficult coding tasks.
Knowledge work leadership: The model outperformed competitors on Hebbia’s Finance Benchmark for senior-level reasoning and aced trading-analysis evaluations by IMC.
Life sciences breakthrough: Mythos 5 has accelerated drug design processes by approximately tenfold and consistently produced novel scientific hypotheses in molecular biology and genomics research. These results suggest the model could meaningfully impact how researchers approach complex biological problems.
What Comes Next
Anthropic’s dual launch underscores a deliberate strategy to expand access to advanced AI while maintaining focus on safety. The company aims to continuously improve its safeguards as more capable models emerge. This approach seeks to maximize potential for societal good, as demonstrated by the models’ contributions to securing critical software and accelerating life sciences research.
The real test will be how well these safeguards hold up in practice as Fable 5 reaches more users and more creative use cases emerge in the wild.
Follow Hashlytics on Bluesky, LinkedIn, Telegram and X to Get Instant Updates
