BloomBench - Arabic - Create: leaderboard
BloomBench (Almieyar-Oryx series, QCRI, 2026) is a bilingual English and Arabic multiple-choice benchmark for vision-language models organized by the six levels of the revised Bloom taxonomy, scored as zero-shot accuracy in percent from the answer letter parsed from generated text (chance 25). This board is accuracy on the 685 Modern Standard Arabic items of the Create level (picking the synthesis that best satisfies creative constraints); higher is better.
Metric: Accuracy (%) on the 685 Arabic BloomBench items at the Bloom level Create (choosing the synthesis that best meets creative constraints), four-option image-question multiple choice, zero-shot, temperature 0, the chosen letter parsed from the generated answer (Regex-based Answer Extraction; unparseable answers count as wrong); chance 25; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B | 82 | #394 |
| 2 | Gemma 3 27B | 78 | #1631 |
| 3 | Gemma 4 26B A4B | 78 | #615 |
| 4 | Gemma 3 12B | 75 | #2000 |
| 5 | GPT-4o Mini | 69 | #1623 |
| 6 | Qwen 2.5 VL 7B | 69 | #1787 |
| 7 | Qwen 2 VL 7B | 67 | #2261 |
| 8 | Gemma 3 4B | 65 | #5064 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=bloombench-arabic-create · How It Works · Data refreshed daily, snapshot 2026-11-13.