GenPuzzle: Benchmarking Visual Reasoning in Image Generation Models
TL;DR
GenPuzzle provides a testbed for measuring progress from image rendering toward visual problem solving, and uses task-specific evaluation protocols that select the automatic judge by measuring agreement with human reference scores.
Abstract
Recent image generation systems increasingly combine multimodal understanding, reasoning, and synthesis, suggesting that they may do more than render plausible scenes. Yet existing evaluations emphasize aesthetics, prompt alignment, compositionality, or text-based answers, leaving unclear whether these systems can solve visual problems and faithfully express solutions in pixels. We introduce GenPuzzle, a benchmark for reasoning-centric image generation. GenPuzzle contains 2,005 problems across 12 tracks, spanning pattern completion, spatial construction, mazes, Sudoku, nonograms, tangrams, board games, matchstick puzzles, orthographic projection, and mathematical visual proof. Each task provides a visual puzzle and requires an image output that preserves the input state while executing a logically valid solution. GenPuzzle uses task-specific evaluation protocols: discrete grid outputs are transcribed and verified programmatically, while visually complex outputs are assessed with tiered, multidimensional, or binary multimodal large language model (MLLM) rubrics. We further select the automatic judge by measuring agreement with human reference scores. Across three frontier generators, the strongest model reaches only 40.57 Macro Overall, revealing frequent failures in logic, geometry, state preservation, and instruction execution. GenPuzzle provides a testbed for measuring progress from image rendering toward visual problem solving.