Baidu Inc. has published a new model of artificial intelligence that articulates his reasoning, in an apparent attempt to take up the momentum against emerging rivals like Deepseek.
The Ernie X1 model of the Internet research leader in China works similarly to Deepseek R1 – which shocked Silicon Valley by offeringcomparable performanceTo the best chatbots in the world has a fraction of their development cost. Baidu’s reasoning model excels in fields such as daily dialogues, complex calculations and logical deduction, he said in astatementSunday.
Baidu has also improved its flagship foundation model to Ernie 4.5. He immediately rendered all levels of his service – including the X1 – free model for his chatbot users, several weeks previously previously.
The company based in Beijing was the first in the Chinese technological sector of the Dollars Billion to launch a chatbot modeled after the Openai chatgpt, but rival chatbots of bytedance Ltd. And Monshot AI quickly took over. Open source models like Qwen d’Alibaba, then Deepseek have acquired greater recognition in the world developer community.
Ernie 4.5 surpasses the latest GPT 4.5 of Openai in generation of text, said Baidu, citing several landmarks of industry.
Baidu said he would make Ernie Ai models from June 30, representing a major strategic change after the rise of Deepseek. He also joined the R1 model in his search engine – his business and butter business.
The arrow generating AI appeared in the baiduDecember quarter resultsVia a 26% leap in cloud income. This increase, motivated by the services provided to developers hunting IT power, was overshadowed by low advertising sales in the middle of China’s economic discomfort.
Baidu concluded aNet-out dealTo acquire the Yy Live Joyy Inc. streaming platform, taking control of $ 2.1 billion has published some $ 1.6 billion that Baidu previously deposited in entire accounts, what itplansInvest in AI and cloud infrastructure.
This story was initially presented on Fortune.com
