model name, source, v1 metric
Google: Gemma 3 27B ,Open Source,95.4022988%
Google: Gemini Flash 2.0,Proprietary Model,97.7011494%
Google: Gemini 2.0 Flash Lite,Proprietary Model,95.4022988%
DeepSeek: R1,Open Source,96.5517241%
Qwen: QwQ 32B,Open Source,94.2528735%