Baidu's Wenxin Yiyan Updates and Developments

TapTechNews on June 28th. At the Baidu WAVESUMMIT deep learning developers' conference held today, Baidu CTO Wang Haifeng announced that the cumulative user scale of Wenxin Yiyan has reached 300 million, and the daily highest number of invocations has reached 500 million times. For comparison, Li Yanhong said in April this year that the number of Wenxin Yiyan users exceeded 200 million, and the average daily API invocations also exceeded 200 million.

Our rapid development this year, including the rapid development of the entire large model, cannot be separated from the support of Fei Jiang. Wang Haifeng said.

At the conference, Wang Haifeng also announced the release of the 4.0 Turbo version of the Wenxin large model. It is introduced that compared to the 4.0 version of the Wenxin large model, the new Turbo version has a faster response speed and stronger retrieval ability.

TapTechNews noticed that Wenxin Yiyan just updated to version 3.1.0 on June 28th, adding the Wenxin large model 4.0 turbo, and at the same time taking offline the plugin function.

Wenxin Yiyan was released on March 16, 2023. After continuous iterations of versions 3.5 and 4.0, there was also news that Baidu will release the next-generation Wenxin large model 5.0 during the Baidu World Conference in 2025. When Li Yanhong released the 4.0 version last October, he also boasted: Wenxin large model is already comparable to GPT-4.

In May this year, Baidu announced that the two main models of the Wenxin large model, ENIRESpeed and ENIRELite, are completely free. These two models are the main models often used by customers in scenarios, and have a very high invocation volume in scenarios such as copywriting creation, chat dialogue, knowledge Q&A, content continuation, etc.