Coding benchmark
GameDevBench leaderboard
Every model the catalog carries a published GameDevBench value for, ranked by that value.
CategoryCoding
MeasurePass@1 on the full task set with 95% confidence intervals
Tasks333 tasks from 88 tutorials
DifficultyMultimodal game development in Godot 4.4.1
Evaluates coding agents on 333 multimodal game-development tasks in Godot, spanning 2D graphics, 3D graphics, user interfaces, and gameplay logic.
No model in the catalog has a published GameDevBench score.
The benchmark is defined by GameDevBench: Evaluating Agentic Capabilities Through Game Development, but the catalog carries no value for it yet. An absent value is shown as absent here rather than as a zero.