Devin adds Claude Fable 5.2, outscoring Anthropic's Opus 5.5 on coding benchmark

Cognition added Claude Fable 5.2 to Devin Desktop and Devin CLI on November 9, 2026, and published benchmark scores showing it ahead of Anthropic's Opus 5.5 on real-world coding tasks.

Cognition announced on November 9, 2026, that it added Anthropic's new Claude Fable 5.2 model to Devin Desktop and Devin CLI, publishing benchmark results that show the new model outscoring Anthropic's own larger Opus 5.5 model on real-world engineering tasks, according to the source material: Claude Fable 5.2 is now available in Devin.

Benchmark scores on FrontierCode 2.0

Claude Fable 5.2 became available in Devin Desktop and Devin CLI, published November 9, 2026.
On Cognition's FrontierCode 2.0 benchmark, Fable 5.2 scored 49.1%.
Fable 5.1 scored 32.4% on the same benchmark.
Anthropic's Opus 5.5 scored 45.5% on FrontierCode 2.0, lower than Fable 5.2's 49.1%.
FrontierCode 2.0 grades real-world engineering tasks on mergeability and quality.
At low reasoning effort, Fable 5.2 already exceeds Fable 5.1's best recorded score, at less than one third of the per-task cost of that run.

What it means for people who build with AI

For Devin users, the upgrade requires no product changes: Fable 5.2 simply becomes the stronger, cheaper coding model running inside the same Desktop and CLI tools. Cognition's independent FrontierCode 2.0 numbers also give teams a concrete data point when choosing which model to run their coding agents on, since an outside benchmark now shows Fable 5.2 beating Anthropic's own flagship Opus 5.5 on mergeable, real-world engineering output.

Sources

Claude Fable 5.2 is now available in Devin (Cognition (Devin), November 9, 2026, official)

Written with AI assistance, strictly from the sources listed above. Before it went live, automated checks verified its dates, sources and every language edition.