GABench - Graph Theory - Graph Level (Numerical-Attribute Graphs): leaderboard

Metric: Task success rate (%; share of tasks whose final answer a DeepSeek-V4-Flash judge marks consistent with the verified ground truth; graph-level graph theory (centralities, paths, bridges, cycles, diameters and similar structural properties) with its 54 graph-algorithm tools on numerical-attribute graphs (DGraph-Fin, Pokec, NBA, Credit and Bail, 200 generated tasks per dataset and category); LLM agent in the OpenClaw harness, one run). Source: arxiv.org. Saturation forecast: Around March 2028. 6 models tracked.

Top models

#ModelScore
1DeepSeek V4 Pro29.1
2GPT-5 Mini22.54
3GLM-4.5 Flash8.89
4GLM-5-Turbo5.92
5Gemini 3 Flash5.34
6Qwen 3 235B A22B1.82

Interactive version: theaggregate.ai/benchmark?slug=gabench-graph-theory-graph-level-numerical-attribute-graphs · How It Works · Data refreshed daily, snapshot 2026-09-29.