SAHARA - Text Classification Cluster: leaderboard
Metric: Cluster score (mean of task scores). Source: huggingface.co. 47 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 52.67 |
| 2 | GPT-5 | 48.19 |
| 3 | Gemini 2.5 Pro | 48.18 |
| 4 | GPT-4.1 | 48.07 |
| 5 | GPT-5.1 | 47.77 |
| 6 | Claude Sonnet 4 (20250514) | 47.28 |
| 7 | GPT-5.6 Luna | 46.45 |
| 8 | Claude Sonnet 4 | 45.15 |
| 9 | Gemma 3 27B (IT) | 44.44 |
| 10 | Gemini 2.5 Flash | 43.95 |
| 11 | Qwen 3.5 27B | 43.64 |
| 12 | Qwen 3.6 27B | 43.2 |
| 13 | Qwen 3.6 35B A3B | 40.43 |
| 14 | Llama 4 Scout Instruct | 39.02 |
| 15 | c4ai-command-a-03-2025 | 38.64 |
Interactive version: theaggregate.ai/benchmark?slug=sahara-text-classification-cluster · How It Works · Data refreshed daily, snapshot 2026-09-19.