AI in practice

Baidu's Chinese ChatGPT "Ernie Bot" reaches 70 million users in three months

Matthias Bastian
The Chinese search and Internet company Baidu presents the latest version of its chatbot Ernie, which is supposed to deliver significantly better results than before. It also gets support for plugins.

Midjourney prompted by THE DECODER

With Ernie Bot, the Chinese Internet company Baidu has probably the strongest Chinese ChatGPT competitor on the market. Now the company announced its first numbers.

China's leading internet search company Baidu has beaten financial expectations in the last quarter, thanks in part to generative AI. The company's AI chatbot, Ernie Bot, China's answer to OpenAI's ChatGPT, has reached 70 million users in the three months since its launch.

Ernie Bot is integrated into all of Baidu's major products, and Baidu charges a monthly subscription fee of about eight US dollars to use it.

But Ernie Bot is nowhere near the numbers of OpenAI's ChatGPT: After a dip in the summer, ChatGPT's web traffic rebounded to about 1.7 billion visits worldwide in October 2023, close to its peak of 1.8 billion in May 2023.

In the U.S., traffic increased to 164.3 million visits in September and could exceed 195 million visits in October, which would be a record.

AI model flood in China

Baidu founder Robin Li recently expressed concern about China's "war of a hundred models," in which tech giants and venture capitalists invest billions of dollars in startups that develop new AI models from scratch, often on an open-source basis.

Li believes this is a waste: the goal should be AI applications with millions of users, not a multitude of AI models. According to Li, the AI model on which the Ernie bot is based matches the capabilities of OpenAI's GPT-4.

Rong Luo, Baidu's chief financial officer, said in the company's quarterly earnings announcement that the company will continue to prioritize investment in generative AI and basic models, while relentlessly focusing on efficiency and strategic resource allocation.

Western technology companies also face the challenge of controlling the cost of generative AI as they scale.