HELM Arabic - Alghafa - Multiple Choice Sentiment Task: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Haiku 4.5 (20251001) | 46.1 |
| 2 | Qwen 3.5 397B A17B | 45.5 |
| 3 | GPT-4.1 (2025-04-14) | 43.9 |
| 4 | Llama 4 Scout Instruct | 43.7 |
| 5 | DeepSeek V3.1 | 43.7 |
| 6 | Gemini 2.5 Flash (Thinking) | 43.7 |
| 7 | Llama 3.3 70B Instruct | 43.3 |
| 8 | GPT-5.4 (2026-03-05) | 43.1 |
| 9 | Qwen 2.5 72B Instruct | 43 |
| 10 | Claude Opus 4.7 | 43 |
| 11 | Llama 4 Maverick Instruct FP8 | 42.7 |
| 12 | Gemini 2.5 Flash Lite (Thinking) | 42.2 |
| 13 | Claude Sonnet 4.6 | 42.1 |
| 14 | AceGPT-v2-32B-Chat | 42.1 |
| 15 | Mistral Large 3 | 41.9 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-alghafa-multiple-choice-sentiment-task · How It Works · Data refreshed daily, snapshot 2026-09-19.