Devin Fusion hits 61.7 on Coding Agent Index, costs 36% less than Claude Code per task

ArtificialAnlys · x · 2026-09-12

Artificial Analysis published detailed per-eval results for both Devin Fusion configurations.

Devin Fusion CLI with Claude Fable 5.1 (xhigh) + SWE-2 (medium) scores 61.7 on the Coding Agent Index v1.5, nearly tied with Claude Code's Fable 5.1 (max, with fallback) at 62.2, while costing 36% less ($7.9 vs $12.4 per task) with essentially flat speed (35.8 vs 34.8 min/task). Per-eval: DeepSWE 1.1 63.1 vs 64.3, SWE-Atlas QnA 65.9 vs 64.8, Terminal-Bench 4.0 56.1 vs 57.6. Both configs sit on the Pareto frontier of score vs cost.

Original post →

More from coding & agent

coding & agent channel →