SciEvalKit: leaderboard
Scientific capability ranking from OpenCompass and SciEvalKit, measuring frontier models on science-heavy reasoning and knowledge tasks.
Metric: Scientific Capability Score. Source: github.com. Saturation forecast: Around June 2028. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 48.74 | #80 |
| 2 | Claude Opus 4.5 | 45.78 | #83 |
| 3 | Qwen 3 Max | 45.33 | #208 |
| 4 | GPT-5 | 45.12 | #98 |
| 5 | Claude Sonnet 4.5 | 42.12 | #142 |
| 6 | Seed 1.8 | 42.05 | #139 |
| 7 | Qwen 3 VL 235B A22B | 41.43 | #238 |
| 8 | DeepSeek V3.2 | 41.35 | #205 |
| 9 | O3 | 41.08 | #124 |
| 10 | Gemini 2.5 Pro | 40.77 | #152 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=scievalkit · How It Works · Data refreshed daily, snapshot 2026-10-20.