Benchmark
ScreenSpot-v2 - Web
Vision benchmarks measure how well models understand images. Scores are percentages.
- Models measured
- 5
- Scores recorded
- 5
- Sources
- 1
- IQ weight
- 0.37
Score distribution
- 50.6%
- Mean
- 39.1%
- Median
- 32.9%
- Std dev
- 12.4%
- 10th percentile
- 93.4%
- 90th percentile
- 12.4%
- Min
This benchmark has a weight of 0.37 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Holo2 30B A3B· ScreenSpot data | 93.4% |
| 2 | Qwen2.5-VL 7B Instruct· ScreenSpot data | 85.3% |
| 3 | qwen2.5-vl-3b-instruct· ScreenSpot data | 39.1% |
| 4 | Qwen2-VL-72B· ScreenSpot data | 22.9% |
| 5 | GPT 4o· ScreenSpot data | 12.4% |
Data credits: ScreenSpot grounding leaderboard (GUI-Agent), Apache-2.0 (licence)