Gemini 3 Flash: benchmark results
Google's speed-oriented Gemini 3 Flash model, balancing capability with lower latency. Provider: Google. Released 2025-12-17. Access: API.
Aggregate Rating 1763 ± 2, rank #99 of 1656 rated models, from 2161 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| A Matter of TASTE | 82 | Airline (Pass@1) τBV (self-reported) | 100 |
| AGGBench | 0.73 | Normalized Absolute Count Error (mean; |predicted count - go | 100 |
| AGGBench - Evidence Recall | 49 | Chunk-level evidence recall (%; share of gold evidence chunk | 100 |
| ASCIIArt-Bench - Category Selection | 38.5 | Accuracy (%) on ASCIIArt-Bench's 200 ASCII-art understanding | 100 |
| ASCIIArt-Bench - Direct Understanding | 30.5 | Accuracy (%) on ASCIIArt-Bench's 200 ASCII-art understanding | 100 |
| Ambig-DS - Objective (Ask) | 31 | Normalized task score (0-100) in the Ask condition (before m | 100 |
| Ambig-DS - Target (Ask) | 61 | Normalized task score (0-100) in the Ask condition (before m | 100 |
| AndroidDaily | 62 | Task success rate (%) over all 350 tasks in 94 real closed-s | 100 |
| AndroidDaily - 3+ Constraints | 58 | Task success rate (%) on tasks with three or more constraint | 100 |
| AndroidDaily - Creation and Communication | 61.6 | Task success rate (%) on content creation and communication | 100 |
| AndroidDaily - Cross-App | 59.5 | Task success rate (%) on tasks spanning two or more apps, 35 | 100 |
| AndroidDaily - Information and Decision | 63.6 | Task success rate (%) on information-seeking and decision ta | 100 |
Interactive version: theaggregate.ai/model?slug=gemini-3-flash · How It Works · Data refreshed daily, snapshot 2026-10-20.