Baidu has unveiled its AI model, Ernie X1, which is designed to articulate its reasoning process. The Ernie X1 model operates on similar principles to DeepSeek R1, a model that garnered significant attention in Silicon Valley for its ability to deliver comparable performance to leading chatbots at a significantly lower development cost. Baidu emphasises that its reasoning model excels in areas such as daily dialogue, complex calculations, and logical deduction, as per a report.
Bloomberg reports that in addition to the Ernie X1 launch, China’s leading internet search provider has upgraded its flagship foundation model to Ernie 4.5. Notably, the company has made all tiers of its AI service, including the X1 model, freely accessible to its chatbot users, accelerating its initial plans by several weeks.
Baidu says Ernie 4.5 outperforms OpenAI’s GPT 4.5
Baidu claims that Ernie 4.5 outperforms OpenAI’s latest GPT 4.5 in text generation, citing various industry benchmarks. Baidu has also announced its intention to make its Ernie AI models open-source starting June 30. The company has also integrated the R1 model into its core search engine business.
This launch appears to be a strategic move to regain momentum in the competitive AI landscape, where Baidu faces increasing pressure from emerging rivals such as DeepSeek, the report added.
While Baidu was the first major Chinese tech company to introduce a ChatGPT-like chatbot, it has since faced stiff competition from rivals like TikTok parent company ByteDance and Moonshot AI. Furthermore, open-source models such as Alibaba’s Qwen and DeepSeek have received recognition within the global developer community, the report added.
The generative AI boom has positively impacted Baidu’s financial performance, with a 26% increase in cloud revenue reported in its December-quarter results. This growth, driven by services provided to developers seeking computing power, was somewhat offset by weaker advertising sales amidst China’s challenging economic conditions.