Recently, the FlagEval weighing model evaluation platform created by Beijing Zhiyuan Research Institute has been comprehensively upgraded, and the ranking list of FlagEval models for 202406 has been released. The latest ranking shows that the Baidu Wenzhen 4.0 model ranks first with a comprehensive score of 89.72 among closed-source dialogue models, surpassing many domestic large models such as Byte Cloud, Douyinbao, Ali Tongyi Qianwen, as well as OpenAI's latest model GPT-4o. In the context of the Chinese language, the overall performance of domestic leading language models represented by the Wenzhen model has exceeded the performance of international first-class standards.
智源大模型权威评测：百度文心4.0登顶国内外“百模”榜，豆包、kimi等追随其后
Authorized evaluation of Zhixuan Big Model: Baidu Wensin 4.0 topped the domestic and foreign 'Baimo' list, followed by Dou Bao, Kimi, and others.
The translation is provided by third-party software.
The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.