Anthropic launches Claude Fable 5.1 and Mythos 5.1: cheaper and better?

Anthropic launched Claude Fable 5.1 and Mythos 5.1, claiming the models are the world's most advanced for coding and knowledge work, with Fable 5.1 generally available and Mythos 5.1 under controlled access. Fable 5.1 achieves higher benchmark scores than its predecessors and cuts costs by 25% for typical workloads and 45% for agentic tasks, while Mythos 5.1 demonstrates scientific capabilities such as designing protein binders with ten times higher affinity and running ML models up to 2.5 times faster with 30-60% lower GPU costs.

Anthropic introduces Claude Fable 5.1 and Mythos 5.1, two new AI models with improved performance in coding, knowledge work, and scientific research. Anthropic announces Claude Fable 5.1 and Mythos 5.1. Both models are based on the same technology but offer different levels of security and access, depending on the intended use. Claude https://x.com/claudeai/status/2094848572143407483 even calls these models “the world’s most advanced models for coding and knowledge work.” Fable 5.1 is generally available, while Mythos 5.1 is available through a controlled access program for specific sectors such as cybersecurity and the life sciences. The previous version, Mythos 5, is also only accessible to a select audience due to the risk of misuse. A copy of that, Fable 5, is accessible to the general public, although Anthropic imposed some limitations https://itdaily.com/news/security/anthropic-is-already-adjusting-claude-fable-5-restrictions/ on the model. These have since been adjusted. Better and cheaper? Claude Fable 5.1 and Mythos 5.1 show significant progress over their predecessors, especially in coding and complex problem-solving. In benchmarks such as Terminal-Bench-Science, CursorBench, and OSWorld 2.0, Fable 5.1 achieves higher scores than Fable 5 and Opus 5. Fable 5.1 also offers significant cost savings: for typical workloads, costs are approximately 25 percent lower, and for agentic tasks, this increases to 45 percent. Recent figures show that the previous version, Fable 5, attracts only eleven percent of its users and cannot compete with cheaper models https://itdaily.com/blogs/business/anthropic-fable-5-adoption-restricted/ such as Opus 4.8 and Opus 5 https://itdaily.com/news/cloud/anthropic-introduces-claude-opus-5-with-improved-performance-and-lower-costs/ . The focus on a cheaper new version is likely no coincidence. For scientists The scientific capabilities of Claude Mythos 5.1 are highlighted by performance in molecular design and biological modeling. In collaboration with external organizations, it was demonstrated that Mythos 5.1 can design protein binders with a ten times higher affinity than conventional designs. The AI also showed impressive results in computational analyses, such as creating a new elevation map of Venus. Furthermore, the model accelerates research in biology by running machine learning models up to 2.5 times faster and reducing GPU costs by 30 to 60 percent. Security Anthropic has paid extra attention to the safety, security, and alignment of the models. According to the company, Fable 5.1 and Mythos 5.1 underwent extensive testing regarding chemical and biological risks, cyber threats, and behavior in agentic scenarios.