According to media reports, insiders say that Baidu will release the latest version of the Wenxin big model inference model as early as the end of August, and the model is currently undergoing full testing. In terms of logical reasoning, complex calculations, and other abilities, the new version of Wenxin's reasoning model will surpass OpenAI O3's full blooded version. In addition, Baidu also plans to release the latest version of the Wenxin pedestal model in the coming months. An insider pointed out that the new model is likely to be released at the Baidu World Conference.
According to the latest news, OpenAI officially released the GPT-5 model during a live broadcast early this morning (8th), claiming to be its most outstanding model with top-notch performance in programming, mathematics, writing, health, visual perception, and other fields.
Currently, global artificial intelligence companies are also competing to release new models, and improving model inference capabilities has become a core competitive indicator for model companies. As the first technology giant in China to launch ChatGPT like chatbots, Baidu's upgrade plan is timely and aims to compete with companies such as OpenAl and DeepSeek. As chatbots shift from simple conversations to complex tasks such as programming, planning, and problem-solving, reasoning ability is becoming increasingly important.
