Anthropic ships Fable 5.1 and Mythos 5.1, cuts cache reads 75%

Anthropic released Claude Fable 5.1 and Mythos 5.1 on September 1 at unchanged API prices: $10 per million input tokens and $50 per million output tokens. Cache reads are 75% cheaper, which the announcement says is "cutting typical workload costs by roughly 25% and highly agentic workloads by up to 45%."

The two are one underlying model shipped under different safeguard regimes. Fable 5.1 is generally available; Mythos 5.1 is restricted to vetted cybersecurity and life-sciences organizations that need the "same model, but with different levels of safeguards."

The headline benchmark is a doubling. According to Decrypt, Fable 5.1 scored 52.6% on Terminal-Bench-Science 0.1 against Fable 5's 24.7%, and 55.8% on Terminal-Bench 4.0 against 42.0%. On Humanity's Last Exam it reached 60.9% without tools and 65.0% with tools.

The safeguard reset is quieter but bigger. Anthropic reports that the new cybersecurity safeguards block "60% fewer false positives than before," and that biology safeguards now fire 85% less often on benign biology and medical questions than the ones that launched with Fable 5. Fable 5.1 is now permitted to discover software vulnerabilities, though not to develop exploits for them.

The company also credits the model with concrete science outputs: Mythos 5.1 "designed experimentally validated protein binders," and Fable 5.1 "trained a neural network that produced a higher-resolution elevation map covering roughly a third of Venus," per VentureBeat. Three of the researchers we track shared the announcement the day it went live.

Fable 5.1 is available via the Claude API and on Amazon Bedrock, Google Cloud, and Microsoft Foundry under the model ID "claude-fable-5-1."