Kaggle WWTP Engineering: leaderboard
Metric: Score (%). Source: www.kaggle.com. 35 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 84.62 |
| 2 | Claude Sonnet 4 | 84.62 |
| 3 | DeepSeek V3.2 | 78.57 |
| 4 | GPT-5.4 | 76.92 |
| 5 | Claude Haiku 4.5 | 76.92 |
| 6 | Gemini 3.1 Flash Lite (Preview) | 76.92 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 76.92 |
| 8 | Claude Opus 4.5 | 76.92 |
| 9 | GPT-5.4 Mini | 76.92 |
| 10 | Gemini 2.0 Flash Lite | 76.92 |
| 11 | Gemini 3.1 Pro (Preview) | 71.43 |
| 12 | Claude Opus 4.7 | 71.43 |
| 13 | Gemini 2.5 Pro | 69.23 |
| 14 | Gemini 2.5 Flash | 69.23 |
| 15 | Gemini 2.0 Flash | 69.23 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-wwtp-engineering · How It Works · Data refreshed daily, snapshot 2026-09-19.