Machine learning on graphs has made substantial progress across domains such as molecular property prediction and chip design. Yet benchmarking practices remain fragmented, often relying on narrow, task-specific datasets and inconsistent evaluation protocols, hindering reproducibility and broader progress. With the recent popularity of graph foundation models, these weaknesses have become apparent, as existing benchmarks are insufficient for thorough evaluation. To address these challenges, we introduce GraphBench, a comprehensive benchmark suite spanning diverse real-world domains and task settings, including node-level, edge-level, graph-level, and generative tasks. GraphBench provides standardized evaluation protocols, including consistent dataset splits and metrics for assessing out-of-distribution generalization across selected tasks, as well as a unified hyperparameter-tuning framework. We further evaluate GraphBench with recent message-passing neural networks and graph transformer models, establishing principled baselines for future research. See www.graphbench.io for further details.
@article{arxiv.2512.04475,
title = {GraphBench: Next-generation graph learning benchmarking},
author = {Timo Stoll and Chendi Qian and Ben Finkelshtein and Ali Parviz and Darius Weber and Fabrizio Frasca and Hadar Shavit and Antoine Siraudin and Arman Mielke and Marie Anastacio and Erik Müller and Maya Bechler-Speicher and Michael Bronstein and Mikhail Galkin and Holger Hoos and Mathias Niepert and Bryan Perozzi and Jan Tönshoff and Christopher Morris},
journal= {arXiv preprint arXiv:2512.04475},
year = {2026}
}