Gemini 3 Flash (Preview) (Medium): benchmark results
Gemini 3 Flash (Preview) evaluated at the medium reasoning-effort setting. Provider: Google. Released 2025-12-17. Access: API.
Aggregate Rating 1766 ± 17, rank #205 of 2168 rated models, from 46 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Alexandria - Human Dialectness Rating (Jordanian Arabic) | 4.87 | Dialectness and fluency (1-5; native Jordanian speakers rate | 100 |
| Alexandria - Human Dialectness Rating (Lebanese Arabic) | 4.57 | Dialectness and fluency (1-5; native Lebanese speakers rate | 100 |
| Alexandria - Human Dialectness Rating (Libyan Arabic) | 4.05 | Dialectness and fluency (1-5; native Libyan speakers rate wh | 100 |
| Alexandria - Human Dialectness Rating (Moroccan Arabic) | 4.35 | Dialectness and fluency (1-5; native Moroccan speakers rate | 100 |
| Alexandria - Human Dialectness Rating (Omani Arabic) | 4.55 | Dialectness and fluency (1-5; native Omani speakers rate whe | 100 |
| Alexandria - Human Dialectness Rating (Saudi Arabic) | 4.63 | Dialectness and fluency (1-5; native Saudi speakers rate whe | 100 |
| Alexandria - Human Dialectness Rating (Sudanese Arabic) | 4.77 | Dialectness and fluency (1-5; native Sudanese speakers rate | 100 |
| Alexandria - Human Dialectness Rating (Syrian Arabic) | 3.92 | Dialectness and fluency (1-5; native Syrian speakers rate wh | 100 |
| Alexandria - Human Dialectness Rating (Tunisian Arabic) | 3.24 | Dialectness and fluency (1-5; native Tunisian speakers rate | 100 |
| Alexandria - Human Dialectness Rating (Yemeni Arabic) | 4.49 | Dialectness and fluency (1-5; native Yemeni speakers rate wh | 100 |
| BeQu - Experiment 1 - Entailment Recall | 41.4 | Entailment Recall (%) | 100 |
| IKP - T7 | 28.49 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=gemini-3-flash-preview-medium · How It Works · Data refreshed daily, snapshot 2026-10-20.