TextClass Benchmark — leaderboard
TextClass Benchmark evaluates LLMs and transformers for social-science text classification across multiple domains and languages, reporting domain-specific Elo leaderboards and a weighted Meta-Elo aggregate.
Metric: Meta-Elo (self-reported). Source: benchmarklist.com. Status: saturation imminent. 108 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o | 1825.22 |
| 2 | Gemini 1.5 Pro | 1782.7 |
| 3 | GPT-4 Turbo | 1781.47 |
| 4 | O1 | 1768.81 |
| 5 | GPT-4.5 | 1767.86 |
| 6 | Grok 2 (1212) | 1758.36 |
| 7 | Llama 3.1 405B | 1755.81 |
| 8 | GPT-4 | 1747.59 |
| 9 | Llama 3.3 70B | 1746.41 |
| 10 | Grok Beta | 1741.94 |
| 11 | DeepSeek V3 | 1732.54 |
| 12 | Mistral Large | 1720.36 |
| 13 | Gemini 2.0 Flash | 1701.95 |
| 14 | Gemini 2.0 Flash Lite | 1687.68 |
| 15 | O3 Mini | 1684.99 |
Interactive version: theaggregate.ai/benchmark?slug=textclass-benchmark · How the rankings work · Data refreshed daily, snapshot 2026-07-22.