LOM Benchmark - TreeCheck: leaderboard
The "TreeCheck" subtask of LOM Benchmark. Yonyou AI Lab benchmark testing LLMs on enterprise knowledge-graph and graph-algorithm reasoning posed in natural language — entity, attribute and relation queries plus classic graph tasks such as cycle detection, shortest path, PageRank...
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 27B | 95 | #187 |
| 2 | Qwen 3 Max | 88 | #208 |
| 3 | Qwen 3 4B Instruct | 88 | #657 |
| 4 | DeepSeek V3.2 | 84 | #205 |
| 5 | Qwen 2.5 32B Instruct | 54 | #498 |
| 6 | Qwen 3.5 2B | 3 | #1148 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=lom-benchmark-treecheck · How It Works · Data refreshed daily, snapshot 2026-10-20.