Benchmark
Sayf AthenaBench VSP normalized CVSS MAD
Vision benchmarks measure how well models understand images. Higher scores are better.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 0.78
Score distribution
- 0.781
- Mean
- 0.857
- Median
- 0.107
- Std dev
- 0.716
- 10th percentile
- 0.887
- 90th percentile
- 0.589
- Min
This benchmark has a weight of 0.78 for chat models, measured over 6 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 | Claude Sonnet 4.6athena_vsp· Sayf cybersecurity evaluations data docs | 0.887 |
| 2 | Gemma 4 31B ITathena_vsp· Sayf cybersecurity evaluations data docs | 0.878 |
| 3 | GPT 5.4athena_vsp· Sayf cybersecurity evaluations data docs | 0.857 |
| 4 | GPT OSS 20bathena_vsp· Sayf cybersecurity evaluations data docs | 0.757 |
| 5 | Llama 3.3 70B Instructathena_vsp· Sayf cybersecurity evaluations data docs | 0.716 |
| 6 | Qwen3.6 35B A3Bathena_vsp· Sayf cybersecurity evaluations data docs | 0.589 |
Data credits: Sayf-eval. Copyright (c) 2026 Qatar Computing Research Institute (QCRI). MIT License. (licence)