HELM TORR - Qtsumm: leaderboard
The "Qtsumm" subtask of HELM TORR. Stanford HELM's TORR project (Table-Oriented Reasoning and Retrieval): question answering and reasoning over tables and semi-structured records, including financial filings, scored on program and answer accuracy.
Metric: ROUGE-L. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o (2024-11-20) | 43.38 | #378 |
| 2 | Qwen 2 72B Instruct | 42.07 | #579 |
| 3 | Llama 3.1 70B Instruct | 41.77 | #566 |
| 4 | Llama 3.1 405B Instruct | 41.61 | #461 |
| 5 | Mixtral 8x22B Instruct | 41.3 | #733 |
| 6 | DeepSeek V3 | 41 | #323 |
| 7 | GPT-4o Mini (2024-07-18) | 39.74 | #561 |
| 8 | Mistral 7B Instruct (v0.3) | 38.05 | #1276 |
| 9 | Gemini 1.5 Pro (002) | 37.9 | #350 |
| 10 | Claude 3.5 Sonnet (20241022) | 36.61 | #291 |
| 11 | Mixtral 8x7B Instruct | 36.12 | #1172 |
| 12 | Llama 3.1 8B Instruct | 34.7 | #1040 |
| 13 | Claude 3.5 Haiku (20241022) | 34.41 | #591 |
| 14 | Gemini 1.5 Flash (002) | 34.12 | #585 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-torr-qtsumm · How It Works · Data refreshed daily, snapshot 2026-10-23.