HyperGVL - Hyperedge Counting: leaderboard

HyperGVL (2026) hypergraph task for vision-language models: counting the hyperedges of a hypergraph, with each of 200 problems posed in 35 textual and visual representations. Accuracy (%); higher is better.

Metric: Accuracy (%) on counting the hyperedges of a hypergraph, 7,000 questions per task (200 problems on synthetic and real-world hypergraphs, each posed in 35 textual, visual and combined hypergraph representations), accuracy (%) averaged over the representations, temperature 0.8, top-p 0.95; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Flash98.31#180
2GPT-4o80.29#754
3Qwen 3 VL 8B59.41#927
4Gemma 3 12B57.91#2091
5InternVL3-8B47.93#1795
6Qwen 2.5 VL 7B Instruct41.94#1919
7GLM-4V-9B39.37#3164
8InternVL2.5-8B22.93#2393
9Molmo-7B-D-092417.94#7206
10DeepSeek-VL217.73#3056
11LLaVA-NeXT-7B11.17#5045
12LLaVA-v1.5-7B8.63#7642

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=hypergvl-hyperedge-counting · How It Works · Data refreshed daily, snapshot 2026-11-22.