Benchmark
VCR
Vision benchmarks measure how well models understand images. Scores are percentages.
- Models measured
- 5
- Scores recorded
- 5
- Sources
- 1
- IQ weight
- 1.36
Score distribution
- 17.8%
- Mean
- 21.9%
- Median
- 5.9%
- Std dev
- 9.3%
- 10th percentile
- 23.2%
- 90th percentile
- 9.3%
- Min
This benchmark has a weight of 1.36 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 | Gemini 1.5 Pro· OpenVLM Leaderboard data | 23.2% |
| 2 | InternVL2 26B· OpenVLM Leaderboard data | 22.7% |
| 3 | MiniCPM-V 2.6· OpenVLM Leaderboard data | 21.9% |
| 4 | InternVL2 4B· OpenVLM Leaderboard data | 12.1% |
| 5 | Phi-3.5 Vision Instruct· OpenVLM Leaderboard data | 9.3% |
Data credits: OpenVLM Leaderboard (OpenCompass / VLMEvalKit), Apache-2.0 (licence)