Anthropic's Claude Fable 5.1 promises better coding and research at up to 45 percent less
decoder ·
Anthropic released Claude Fable 5.1 and Mythos 5.1, claiming better coding and research performance while cutting costs by up to 45 percent for agentic workloads through cheaper cache reads. Fable 5.1 outperforms its predecessor and rival GPT-5.6 Sol on agentic benchmarks. However, Artificial Analysis disputed the savings, noting Fable 5.1 at max effort costs about 20 percent more per task than Fable 5 due to higher output token usage. The two models share a base model but differ in safety guardrails, with Mythos restricted to special access programs. They are also the first Claude models with built-in watermarks, and Anthropic launched a detection API in private preview. Safety filters now produce fewer false positives for cybersecurity and biology queries, and Fable 5.1 can identify software vulnerabilities but not develop exploits. The models are available immediately via API and major cloud platforms.
decoder