China’s Baidu Launches Two New AI Models Amidst Fierce Competition
HONG KONG (Reuters) – Baidu, a leading Chinese technology company, announced on Sunday the launch of two new artificial intelligence models, intensifying the competition in the rapidly evolving AI landscape. One of the new models is designed with enhanced reasoning capabilities, positioning Baidu to compete with other major players.
Baidu stated that one of the models, ERNIE X1, “delivers performance on par with DeepSeek R1 at only half the price.” The company emphasized the X1’s advanced capabilities, including “stronger understanding, planning, reflection, and evolution capabilities,” and highlighting it as the first deep thinking model that autonomously utilizes tools.
DeepSeek, a Chinese AI startup, has been making waves in the industry with its AI models, which it claims are comparable to or even surpass those of leading U.S. companies, but at a lower cost. This has injected new energy into the global AI race.
Baidu’s latest foundation model, ERNIE 4.5, “has excellent multimodal understanding ability,” according to the company. It noted that the model has improved in its language abilities, including enhanced understanding, generation, logic, and memory capabilities. Baidu also highlighted the new model’s “high EQ,” indicating it is better at understanding online memes and satirical content.
Despite being one of the earliest Chinese tech giants to launch a ChatGPT-style chatbot, Baidu has faced challenges in securing widespread adoption for its Ernie large language model, even though it claims performance on par with OpenAI’s GPT-4, as the AI sector becomes increasingly competitive. Multimodal AI systems are designed to process various data types, including text, video, images, and audio, to convert content across formats.
(Reporting by Farah Master and the Beijing newsroom; Editing by Lincoln Feast.)
