Gemini 4 Argon on Q-Bench Video
66.9%Rank #7 of 10 on Q-Bench Video
- Coverage
- Full (868/868)
- Cost per task
- $0.0021
- Median latency
- 3.9s
- Tokens per task
- 824 in · 44 out
Run setting: 1 fps.
Tokens and latency
Tokens per call
| Token type | Mean | Median | 90th pct. |
|---|---|---|---|
| Input | 824 | 811 | 1,520 |
| of which cached | 0 | 0 | 0 |
| Output | 44 | 6 | 113 |
| of which reasoning | 38 | 0 | 107 |
Over 1,736 calls that returned an answer, including questions asked again after a failed attempt.
Latency
- Median
- 3.9s
- Mean
- 7.9s
- 90th percentile
- 11.4s
- 99th percentile
- 84.8s
Per call, end to end including file upload and processing.