Claude Fable 5.1 vs GPT-5.6 Luna: pricing, Quality, Value, and benchmarks

Buyer access

Pricing, app access, and Conversation Value

• Humanity's Last Exam

  Normalized quality input

  59.1%

  Artificial Analysis Claude Fable 5.1 evaluation | Independent HLE result using adaptive reasoning at maximum effort with Anthropic default safeguard fallback.

• SWE-Bench Pro

  Normalized quality input

  81.2%

  Anthropic Claude Fable 5.1 and Mythos 5.1 system card | Vendor-reported SWE-Bench Pro result using adaptive thinking at maximum effort and default sampling settings.

• Terminal-Bench 2.1

  Agentic terminal task completion

  91.4%

  Artificial Analysis Claude Fable 5.1 analysis | Independent Terminal-Bench 2.1 result using adaptive reasoning at maximum effort with default safeguard fallback.

• GPQA Diamond

  Normalized quality input

Anthropic's newest generally available Fable model leads the current scored evidence, with premium $10/$50 token pricing and safeguard fallback for some flagged requests.

Claude Fable 5.1 combines the strongest current locked-variant HLE and SWE-Bench Pro evidence with a 1M-token context window. Its $10 per million input-token and $50 per million output-token prices make it a specialist choice for difficult coding and knowledge work, while safeguard fallback means integrations should surface the model identity returned by the API.

The fastest, lowest-cost GPT-5.6 option for high-volume workloads.

GPT-5.6 Luna is the fastest and lowest-cost tier in the family. Its accepted HLE and SWE-Bench Pro evidence keep it competitive for high-volume workflows, while difficult tasks should still be tested against Terra or Sol.