ConvRe - Text2Re Hard — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | flan-t5-base | 50.2 |
| 2 | flan-t5-small | 49.5 |
| 3 | GPT-3.5 Turbo (0301) | 39 |
| 4 | text-davinci-003 | 34.8 |
| 5 | flan-t5-large | 29.6 |
| 6 | Claude Instant 1.1 | 26.2 |
| 7 | flan-t5-xl | 17.8 |
| 8 | GPT-4 (0314) | 17.1 |
| 9 | flan-t5-xxl | 4.8 |
Interactive version: theaggregate.ai/benchmark?slug=convre-text2re-hard · How the rankings work · Data refreshed daily, snapshot 2026-07-22.