For Developers/Models/Claude Fable 5.1
Closed SourceAnthropicReleased 2026-09

Claude Fable 5.1

Point release of Anthropic's flagship, tuned for long-horizon agentic work with a large effective price cut.

Context window

1M

Input / 1M tokens

$10.00

Output / 1M tokens

$50.00

Provider

Anthropic

List price unchanged from Fable 5, but Anthropic states typical workloads cost 25% less and highly agentic workloads up to 45% less thanks to token efficiency; cache reads dropped 75% to $0.25 per 1M tokens. · Data verified 2026-09-05

Claude Fable 5.1 (September 2026) is the first point release of the Mythos-class tier. The same weights ship as Claude Mythos 5.1 with reduced restrictions for vetted cybersecurity and life sciences professionals. The release targets agentic work: Terminal-Bench-Science more than doubled versus Fable 5 while pure reasoning barely moved, and Anthropic reports 60% fewer false positives on cybersecurity tasks plus stronger prompt-injection resistance. Enterprise Frontier Safeguards enable zero-data-retention deployment on customer-controlled cloud infrastructure.

How to access it

Generally available on the Claude API and all major cloud platforms (AWS, Google Cloud, Microsoft Azure) since September 2026 under the model ID claude-fable-5-1. The reduced-restriction Mythos 5.1 variant requires enrollment in Anthropic's Cyber or Life Sciences Verification Programs.

Strengths

  • Agentic scientific research: Terminal-Bench-Science 0.1 at 52.6% vs 24.7% for Fable 5
  • Agentic coding: Terminal-Bench 4.0 at 55.8% vs 42.0% for Fable 5
  • 25-45% lower effective cost than Fable 5 at the same list price
  • Cache reads at $0.25 per 1M tokens (75% cheaper), compounding on agent loops
  • 60% fewer cybersecurity false positives; improved prompt-injection robustness

Best for developers who...

Long-running agentic pipelines where cache-read cost dominatesAutomated scientific research and computational biology workflowsDefensive security work (via the Cyber Verification Program for Mythos 5.1)Teams that priced out Fable 5 and want to re-check the math

When to choose it (and when not to)

Reach for Claude Fable 5.1 when...

  • You run long agentic workloads and want the 25-45% effective cost reduction
  • Your work is agentic science, biotech, or security tooling - the areas this release moved most
  • You need zero-data-retention enterprise deployment on your own cloud

Benchmarks

BenchmarkScoreNotes
Terminal-Bench-Science 0.152.6%vs 24.7% for Fable 5
Terminal-Bench 4.055.8%vs 42.0% for Fable 5
Humanity's Last Exam (with tools)65.0%vs 63.8% for Fable 5

Tools powered by Claude Fable 5.1

Compare Claude Fable 5.1

Compare Claude Fable 5.1 with any other model

Build a comparison →
All model comparisons →