Overview
- Anthropic announced two 5.1 models that share a base: Fable 5.1 is broadly available through the Claude API and on AWS, Google Cloud, and Azure, while Mythos 5.1 is limited to vetted U.S. organizations under a trusted‑access program.
- The company published internal benchmarks showing Fable 5.1 outperforming its prior models and, by its measures, outperforming OpenAI’s GPT‑5.6 Sol on agentic, programming, and knowledge tests.
- Fable 5.1 cuts cache‑read costs by 75% to $0.25 per million tokens, which Anthropic says lowers typical workload costs by about 25% and cuts highly agentic workload costs by up to 45%.
- Anthropic added enterprise safety and compliance measures: a private‑preview invisible watermark detection API, promises of reduced false positives in security checks, and a phased Enterprise Frontier Safeguards program that offers zero data retention and enhanced protections starting this fall.
- Mythos 5.1 has been shown in company demos to boost molecular‑design and computational biology results but remains gated for vetted researchers and government partners to manage dual‑use biosecurity and cybersecurity risks.