Gemini 3.8 Flash on Perception Test
85.5%Rank #2 of 9 on Perception Test
- Coverage
- Full (454/454)
- Cost per task
- $0.0042
- Median latency
- 8.5s
- Tokens per task
- 2,177 in · 696 out
Run setting: 1 fps.
Tokens and latency
Tokens per call
| Token type | Mean | Median | 90th pct. |
|---|---|---|---|
| Input | 2,177 | 2,072 | 3,237 |
| of which cached | 0 | 0 | 0 |
| Output | 696 | 335 | 1,668 |
| of which reasoning | 694 | 333 | 1,667 |
Over 454 calls that returned an answer, including questions asked again after a failed attempt.
Latency
- Median
- 8.5s
- Mean
- 11.0s
- 90th percentile
- 18.9s
- 99th percentile
- 40.8s
Per call, end to end including file upload and processing.