MMTU - Column Relationship: leaderboard
The "Column Relationship" subtask of MMTU. Massive Multi-Task Table Understanding benchmark with real-world table reasoning tasks including joins, transformations, cleaning, and question answering.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around January 2027. 26 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 | 68.71 | #124 |
| 2 | Grok 3 Mini | 67.88 | #281 |
| 3 | GPT-5 | 64.84 | #98 |
| 4 | Gemini 2.5 Flash | 59.57 | #245 |
| 5 | GPT-5 Mini | 58.53 | #186 |
| 6 | DeepSeek R1 0528 | 58.07 | #222 |
| 7 | Gemini 2.5 Pro | 57.02 | #152 |
| 8 | GPT-OSS-120B | 52.64 | #341 |
| 9 | Qwen 3 32B | 49.6 | #443 |
| 10 | GPT-5 Chat | 45.18 | #241 |
| 11 | Qwen 3 235B A22B 2507 Instruct | 44.23 | #300 |
| 12 | GPT-OSS-20B | 42.41 | #518 |
| 13 | Qwen 3 8B | 40.8 | #674 |
| 14 | Llama 4 Maverick Instruct FP8 | 37.24 | #414 |
| 15 | Llama 3.3 70B | 33.82 | #579 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mmtu-column-relationship · How It Works · Data refreshed daily, snapshot 2026-10-20.