Darija Chatbot Arena — leaderboard
Moroccan Arabic chatbot arena using pairwise human preference battles to rank models on Darija conversation quality.
Metric: Elo Rating. Source: huggingface.co. Status: saturation imminent. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o ChatGPT | 1400.6 |
| 2 | Gemini 1.5 Pro | 1326.1 |
| 3 | Claude 3.5 Sonnet | 1286.6 |
| 4 | GPT-4o | 1264.5 |
| 5 | DeepSeek V3 | 1221.2 |
| 6 | GPT-4o Mini | 1212.6 |
| 7 | Qwen 2.5 72B Instruct | 1188.3 |
| 8 | Llama 3.1 405B | 1155.2 |
| 9 | Command-R+ (08-2024) | 1148.7 |
| 10 | Llama 3.3 70B Instruct | 1136.8 |
| 11 | QwQ 32B-Preview | 1106.4 |
| 12 | Llama 3 8B Instruct | 1099.1 |
| 13 | Grok Beta | 1045.5 |
Interactive version: theaggregate.ai/benchmark?slug=darija-chatbot-arena · How the rankings work · Data refreshed daily, snapshot 2026-07-22.