NGDBench - NGD-Prime Aggregation sMAPE: leaderboard
Metric: Symmetric mean absolute percentage error (0 to 2) between the numeric answer of the model's Cypher query, executed on the perturbed NGD-Prime graph, and the clean-graph answer, for aggregation queries (text-to-Cypher); lower is better. Source: arxiv.org. Saturation forecast: Around January 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.2 | 0.84 | #198 |
| 2 | Qwen 3 Coder 480B A35B Instruct | 0.85 | #302 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=ngdbench-ngd-prime-aggregation-smape · How It Works · Data refreshed daily, snapshot 2026-10-11.