RPCBench - Conditional Strategy Quality: leaderboard
Metric: Strategy Quality (0-100). Source: arxiv.org. Saturation forecast: Around December 2026. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 397B A17B | 87.21 |
| 2 | Qwen 3.5 Plus | 86.91 |
| 3 | DeepSeek V4 Pro | 85.85 |
| 4 | Qwen 3.5 122B A10B | 85.33 |
| 5 | DeepSeek V4 Flash | 84.77 |
| 6 | Claude Sonnet 4.6 | 84.65 |
| 7 | Qwen 3.5 35B A3B | 83.46 |
| 8 | GPT-5.5 | 82.64 |
| 9 | Gemini 3.1 Pro (Preview) | 80.21 |
| 10 | Llama 3.1 70B Instruct | 59.26 |
| 11 | Llama 3.1 8B Instruct | 51.59 |
Interactive version: theaggregate.ai/benchmark?slug=rpcbench-conditional-strategy-quality · How It Works · Data refreshed daily, snapshot 2026-09-24.