Gemini 3.8 Flash (Low): benchmark results
Provider: Google. Released 2026-09-02. Access: API.
Aggregate Rating 1811 ± 13, rank #117 of 2168 rated models, from 109 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| cua-speedrun - OSWorld (50 Tasks) | 91.62 | Task success (%), one run per task in the cua-speedrun Modal | 96.9 |
| Korean CSAT 2026 (Easy Mode) - Society and Culture | 47 | Points (out of 50) | 96.3 |
| Wolfram LLM Benchmarking Project | 69.2 | Correct Functionality (%) | 94.5 |
| AA MMMU-Pro | 84.51 | Accuracy (%) | 93.8 |
| Voice Standard v0.1.3 (gpt-5.6-sol; 30s) - Text Correction | 96 | Published text_correction task_quality (1-100), canonical v0 | 93.6 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 80 | Omniscience Index | 93.5 |
| AA GPQA Diamond | 92.02 | Accuracy (%) | 93.2 |
| AA-Omniscience Index - Law | 30.4 | Omniscience Index | 92.8 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 74 | Omniscience Index | 92.4 |
| AA-Omniscience Accuracy | 52.22 | Accuracy (%) | 91.2 |
| Korean CSAT 2026 (Easy Mode) - Life Science I | 50 | Points (out of 50) | 91.2 |
| Korean CSAT 2026 (Easy Mode) - Chemistry I | 50 | Points (out of 50) | 90.9 |
Interactive version: theaggregate.ai/model?slug=gemini-3-8-flash-low · How It Works · Data refreshed daily, snapshot 2026-10-20.