RPCBench - Domain - Yelp Local - Evidence Distortion: leaderboard
Metric: Evidence Distortion Rate (%). Source: github.com. Saturation forecast: Around December 2026. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 10.23 |
| 2 | Gemini 3.1 Pro (Preview) | 12.21 |
| 3 | Qwen 3.5 122B A10B | 16.26 |
| 4 | Qwen 3.5 397B A17B | 17.25 |
| 5 | Qwen 3.5 Plus | 17.79 |
| 6 | Qwen 3.5 35B A3B | 17.86 |
| 7 | Llama 3.1 8B Instruct | 18.17 |
| 8 | DeepSeek V4 Pro | 22.14 |
| 9 | Claude Sonnet 4.6 | 22.75 |
| 10 | Llama 3.1 70B Instruct | 22.82 |
| 11 | DeepSeek V4 Flash | 23.44 |
Interactive version: theaggregate.ai/benchmark?slug=rpcbench-domain-yelp-local-evidence-distortion · How It Works · Data refreshed daily, snapshot 2026-09-24.