Anthropic Launches Claude Fable 5, Its Most Capable AI Model Yet

Anthropic has released Claude Fable 5, a "Mythos-class" AI model that it describes as the most capable model the company has ever made generally available. The model achieves state-of-the-art results across nearly all benchmark categories, including software engineering, scientific research, knowledge work, and vision-based tasks.

The release also includes Claude Mythos 5, an unrestricted version of the same underlying model designed specifically for cybersecurity defense. Mythos 5 will initially be deployed through Project Glasswing, a collaboration with the US government, where it serves as an upgrade to the earlier Claude Mythos Preview. Anthropic says Mythos 5 has "the strongest cybersecurity capabilities of any model in the world."

Performance and Pricing

Claude Fable 5 demonstrates a significant performance lead over previous models, particularly on long and complex tasks. In early testing, Stripe reported that Fable 5 compressed months of engineering work into days — performing a codebase-wide migration across a 50-million-line Ruby codebase in a single day that would have taken an entire team over two months by hand. On Hebbia's Finance Benchmark for senior-level reasoning, Fable 5 achieved the highest score of any model, with gains in document-based reasoning, chart and table interpretation, and problem solving.

"Fable 5's capabilities exceed those of any model we've ever made generally available. It is state-of-the-art on nearly all tested benchmarks of AI capability." — Anthropic

The model also excels at vision tasks, capable of rebuilding a web app's source code from screenshots alone, and can play complex games like Pokémon FireRed using only raw visual input — a task earlier models required extensive scaffolding to accomplish. Both Fable 5 and Mythos 5 are priced at $10 per million input tokens and $50 per million output tokens, less than half the cost of Claude Mythos Preview.

Safeguards and Access

Anthropic acknowledged the risks of releasing such a capable model. Fable 5 includes safeguards that route queries on certain sensitive topics to Claude Opus 4.8, and Anthropic says these safeguards trigger in less than 5% of sessions. The company is working to reduce false positives as rapidly as possible. For cyberdefenders and infrastructure providers, Mythos 5 lifts some of these restrictions to enable advanced cybersecurity work. Anthropic plans to expand access to Mythos 5 through a broader trusted access program in the coming months.

The launch marks another milestone in the rapidly accelerating frontier of AI capability, as companies race to deliver models that can autonomously handle increasingly complex real-world tasks across software engineering, scientific research, and national security domains.

Source: Anthropic