Gemini 3.8 Flash on EgoSchema
80.8%Rank #3 of 9 on EgoSchema
- Coverage
- Full (500/500)
- Cost per task
- $0.0104
- Median latency
- 16.6s
- Tokens per task
- 11,625 in · 451 out
Run setting: 1 fps.
Tokens and latency
Tokens per call
| Token type | Mean | Median | 90th pct. |
|---|---|---|---|
| Input | 11,625 | 11,521 | 12,030 |
| of which cached | 11,457 | 11,363 | 11,903 |
| Output | 451 | 265 | 986 |
| of which reasoning | 448 | 261 | 982 |
Over 500 calls that returned an answer, including questions asked again after a failed attempt.
Latency
- Median
- 16.6s
- Mean
- 17.7s
- 90th percentile
- 24.4s
- 99th percentile
- 35.5s
Per call, end to end including file upload and processing.