Qwen3.8-Max on Perception Test
83.9%Rank #3 of 9 on Perception Test
- Coverage
- Full (454/454)
- Cost per task
- $0.0398
- Median latency
- 69.6s
- Tokens per task
- 7,250 in · 4,209 out
Run setting: 1 fps native video.
Tokens and latency
Tokens per call
| Token type | Mean | Median | 90th pct. |
|---|---|---|---|
| Input | 7,250 | 7,906 | 10,331 |
| of which cached | 4,236 | 4,608 | 10,240 |
| of which frames or video | 7,154 | 7,834 | 10,246 |
| Output | 4,209 | 3,134 | 10,043 |
Over 486 calls that returned an answer, including questions asked again after a failed attempt.
Latency
- Median
- 69.6s
- Mean
- 94.3s
- 90th percentile
- 217.8s
- 99th percentile
- 312.3s
Per call, end to end including file upload and processing.