SuperCLUE-DSPSafeBench - Overall: leaderboard
Metric: Score. Source: www.superclueai.com. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Yi-1.5-9B-Chat-16K | 79.37 |
| 2 | Phi-3 Mini 4K Instruct | 78.62 |
| 3 | internlm2.5-7B-chat | 77.64 |
| 4 | MiniCPM3-4B | 77.64 |
| 5 | Qwen 2.5 7B Instruct | 73.51 |
| 6 | Mistral 7B Instruct (v0.3) | 73.04 |
| 7 | GLM-4 9B Chat | 72.43 |
| 8 | Gemma 2 9B (IT) | 72.34 |
| 9 | Qwen 2.5 1.5B Instruct | 71.26 |
| 10 | Gemma 2 2B (IT) | 70.42 |
| 11 | Baichuan2-7B-Chat | 65.31 |
| 12 | Llama 3.1 8B Instruct | 61.51 |
Interactive version: theaggregate.ai/benchmark?slug=superclue-dspsafebench-overall · How It Works · Data refreshed daily, snapshot 2026-09-19.