Skip to content

Claude Fable 5.2 is now available in Devin | Devin

Devin

Fable 5.2 achieved a 49.1% score on Cognition's FrontierCode 2.0 benchmark, outperforming Claude Opus 5.5 at 45.5% and Fable 5.1 at 32.4%.

Key points

  • At low reasoning effort, Fable 5.2 surpasses Fable 5.1's highest result at under one-third the cost per task.
  • The model is accessible across both Devin Desktop and Devin CLI.

Why it matters

Devin users can now run coding tasks on a higher-scoring model while lowering execution costs compared to previous Fable versions.

Sources