Anthropic Launches Claude Fable 5.1 With Up to 45% Lower Agentic Costs
Synopsis
Anthropic has launched Claude Fable 5.1 with stronger performance, lower agentic costs, improved safeguards and new privacy measures for enterprise customers.
Anthropic has launched Claude Fable 5.1, an upgraded version of its Mythos-class AI model that the company says delivers stronger performance while reducing costs for businesses running long, complex AI tasks.
A restricted version, Claude Mythos 5.1, is also being rolled out to organisations participating in Project Glasswing.
The biggest change for customers is pricing efficiency. Anthropic says Fable 5.1 is typically about 25% cheaper than its predecessor for real-world workloads and can be up to 45% cheaper for highly agentic tasks.
The savings come largely from a sharp reduction in the cost of processing cached data, which allows AI agents to reuse information already processed during lengthy workflows.
Anthropic Targets Cost and Safety Concerns
Fable 5.1 retains a 1 million-token context window and supports up to 128,000 output tokens, making it suited to coding and other knowledge-intensive work. Anthropic has also adjusted its safeguards after customers complained that Fable 5 could be overly restrictive.
The company says Fable 5.1 is now better at distinguishing legitimate biology questions from potentially harmful requests. It can also identify software vulnerabilities, although Anthropic continues to route higher-risk cybersecurity activities, including penetration testing and exploit generation, to its more restricted models.
Privacy is another area receiving attention. Anthropic plans to introduce Enterprise Frontier Safeguards later this year, allowing customer data to be stored on the customer’s own cloud infrastructure rather than Anthropic’s systems.
Mythos 5.1 Remains Restricted
Claude Mythos 5.1 is not receiving a broad public release. It remains limited to vetted Project Glasswing participants working on advanced cybersecurity and related research. Project Glasswing was created to give selected organisations access to Mythos-class capabilities for finding and fixing vulnerabilities in critical software.
Anthropic’s latest release therefore goes beyond a conventional model upgrade. By cutting the cost of persistent AI workloads while loosening some safeguards, the company is trying to make its most capable Claude models more practical for everyday enterprise use without abandoning the controls surrounding higher-risk applications.
Source: The Verge
Vishal is an experienced Editor at Inspirepreneur Magazine with key interests in artificial intelligence, eCommerce, entrepreneurship, lifestyle and startup sector. Prior to joining Inspirepreneur, he was a Content Writer cum Correspondent at Siliconindia Magazine, where he worked on Company Profiles, Cover Stories, Executive Profiles, Feature Articles and Thought Leadership content.
