Is Claude Fable 5.1 available now?

Yes. Anthropic made it generally available on September 1, 2026, on its own platform and through AWS, Google Cloud and Microsoft Azure.

How much does Claude Fable 5.1 cost?

Standard rates are unchanged at $10 per million input tokens and $50 per million output tokens, but cache reads dropped 75% to $0.25 per million tokens — cutting real-world costs by roughly 25–45% depending on the workload.

What is Claude Mythos 5.1 and who can use it?

It is the same underlying model as Fable 5.1 but with lighter safeguards, available only to vetted cybersecurity and life-sciences organizations through Anthropic’s verification programs.

How does Fable 5.1 differ from Fable 5?

Anthropic reports large benchmark gains — including 52.6% vs. 24.7% on Terminal-Bench-Science — plus fewer safety false positives and lower costs on agentic workloads.

Anthropic Launches Claude Fable 5.1 and Mythos 5.1

Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, upgrading its flagship models with stronger coding results and a sharp cut to one of the biggest recurring costs of running AI agents: cache reads.

The two models share the same underlying weights but differ in their safeguards. Fable 5.1 is generally available immediately, both on Anthropic’s own platform and through AWS, Google Cloud and Microsoft Azure. Mythos 5.1 carries lighter restrictions and is reserved for vetted organizations working in cybersecurity or life sciences, distributed only through Anthropic’s Cyber Verification Program and Life Sciences Verification Program.

Same sticker price, cheaper in practice

Anthropic left its headline pricing unchanged, at $10 per million input tokens and $50 per million output tokens. The real change is in cache reads, which dropped 75%, from $1.00 to $0.25 per million tokens. Because agentic coding tools repeatedly re-read large chunks of context, the company says the cut lowers total costs by roughly 25% for typical workloads and up to 45% for heavily agentic tasks.

Benchmark gains

According to Anthropic, Fable 5.1 scored 52.6% on Terminal-Bench-Science 0.1, more than double Fable 5’s 24.7%. On the broader Terminal-Bench 4.0 benchmark, Fable 5.1 reached 55.8% and Mythos 5.1 hit 60.9%, both well above Fable 5’s 42.0%. The company also reported 60.9% on Humanity’s Last Exam without external tools (65.0% with tools) and 73.4% on CursorBench 3.2.0.

Fewer safety false positives

Anthropic said the update also refines its safety systems: cybersecurity-related false positives fell 60%, while biology-related safeguards now trigger 85% less often on benign requests — extending a pattern the company eased for Fable 5 earlier this year. Mythos 5.1 can flag likely software vulnerabilities in code it reviews but is designed to stop short of producing working exploits.

Fable 5.1 succeeds Claude Fable 5, which launched as Anthropic’s flagship model earlier this year, while Mythos 5.1 replaces the restricted-access Mythos 5 tier used for sensitive cybersecurity and biology work.

Same sticker price, cheaper in practice

Benchmark gains

Fewer safety false positives

Frequently asked questions