RPCBench - Error Group - Underspecified Premise - Proactive Detection: leaderboard
Metric: Detection Rate (%). Source: github.com. Saturation forecast: Around 2033. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4 Pro | 15.15 |
| 2 | Claude Sonnet 4.6 | 11.77 |
| 3 | Qwen 3.5 35B A3B | 11.77 |
| 4 | Gemini 3.1 Pro (Preview) | 11.54 |
| 5 | DeepSeek V4 Flash | 11.08 |
| 6 | Qwen 3.5 122B A10B | 11.08 |
| 7 | Qwen 3.5 Plus | 10.77 |
| 8 | Qwen 3.5 397B A17B | 9.77 |
| 9 | GPT-5.5 | 8.38 |
| 10 | Llama 3.1 8B Instruct | 5 |
| 11 | Llama 3.1 70B Instruct | 3.85 |
Interactive version: theaggregate.ai/benchmark?slug=rpcbench-error-group-underspecified-premise-proactive-detection · How It Works · Data refreshed daily, snapshot 2026-09-24.