RPCBench - Error Group - Unverifiable Premise - Conditional Localization: leaderboard
Metric: Localization Accuracy (%). Source: github.com. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 93.64 |
| 2 | DeepSeek V4 Pro | 90.2 |
| 3 | Qwen 3.5 Plus | 89.98 |
| 4 | DeepSeek V4 Flash | 89.81 |
| 5 | Qwen 3.5 122B A10B | 89.62 |
| 6 | Qwen 3.5 397B A17B | 89.49 |
| 7 | Claude Sonnet 4.6 | 88.38 |
| 8 | Qwen 3.5 35B A3B | 87.95 |
| 9 | Gemini 3.1 Pro (Preview) | 83.86 |
| 10 | Llama 3.1 70B Instruct | 51.42 |
| 11 | Llama 3.1 8B Instruct | 50.69 |
Interactive version: theaggregate.ai/benchmark?slug=rpcbench-error-group-unverifiable-premise-conditional-localization · How It Works · Data refreshed daily, snapshot 2026-09-24.