Molmo 2 8B on EgoSchema
58.2%Rank #9 of 9 on EgoSchema
- Coverage
- Full (500/500)
- Cost per task
- —
- Median latency
- 33.7s
- Tokens per task
- 16,280 in · 2 out
Run setting: 1 fps.
Tokens and latency
Tokens per call
| Token type | Mean | Median | 90th pct. |
|---|---|---|---|
| Input | 16,280 | 16,247 | 16,409 |
| of which cached | 157 | 0 | 0 |
| Output | 2 | 2 | 2 |
Over 500 calls that returned an answer, including questions asked again after a failed attempt.
Latency
- Median
- 33.7s
- Mean
- 33.4s
- 90th percentile
- 34.3s
- 99th percentile
- 34.9s
We only consider the first question against a video for self hosted models, ignoring prefix caching latency.