VIALS: leaderboard
Gel blots, microscopy, plasmid maps, flow cytometry plots — the visual artifacts a working life scientist reads before deciding what to do next. Scored as direct vision-language accuracy with no tools, where the best model manages roughly a quarter of the questions.
Metric: Direct VLM Accuracy (%). Source: arxiv.org. Saturation forecast: Around 2030. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 26.5 | #19 |
| 2 | Gemini 3.7 Flash | 26.5 | #23 |
| 3 | Claude Opus 5 | 24.4 | #12 |
| 4 | Gemini 3.1 Pro (Preview) | 21.5 | #58 |
| 5 | Kimi K3 | 20.9 | #30 |
| 6 | Grok 4.6 | 20.5 | #26 |
| 7 | GLM-4.6V | 10.8 | #316 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=vials · How It Works · Data refreshed daily, snapshot 2026-10-20.