Anthropic Unveils "Claude Fable 5.1"... Introduces Mythos Model with Enhanced Safeguards

"Specialized in Multi-Stage Agent Tasks"

Anthropic has launched its next-generation artificial intelligence (AI) models, "Claude Fable 5.1" and "Claude Mythos 5.1."

On September 1 (local time), Anthropic announced the release of Fable 5.1, which specializes in multi-stage agent tasks that encompass coding and knowledge-based work.

Fable 5.1 demonstrated superior performance compared to Fable 5 on key benchmarks, including agent coding, scientific research, multidisciplinary reasoning, and knowledge work.

Cost efficiency for enterprise customers has also been improved. By reducing the cache read price by 75%, costs can be lowered by approximately 25% for general workloads compared to Fable 5, and by up to about 45% in high-load environments such as agent-related tasks.

Other pricing remains the same as Fable 5, at $10 per million input tokens and $50 per million output tokens.

In terms of data security, Anthropic has newly introduced the "Enterprise Frontier Safeguard (EFS)." EFS stores customer data on the customer's own cloud infrastructure rather than on Anthropic's servers, providing a privacy level equivalent to a non-retention policy.

Fable 5.1 can also be used for defensive cybersecurity tasks, including software vulnerability detection.

Mythos 5.1, which was released alongside Fable 5.1, is based on the same foundational model, but is equipped with separate safety mechanisms to support research in cybersecurity and life sciences.

It is provided exclusively to verified institutions through Anthropic’s trust-based access program and is not available to the general public.

This content was produced with the assistance of AI translation services.