The logo of Baidu’s AI chatbot Ernie Bot is displayed near a screen showing the Baidu logo, in this illustration picture taken June 28, 2023. REUTERS/Florence Lo/Illustration/File PhotoThe logo of Baidu’s AI chatbot Ernie Bot is displayed near a screen showing the Baidu logo, in this illustration picture taken June 28, 2023. REUTERS/Florence Lo/Illustration/File PhotoChina’s Baidu (9888.HK) announced on Sunday the launch of two new artificial intelligence models, including a reasoning-focused model designed to compete with DeepSeek’s offerings, as the AI industry becomes increasingly competitive.
Chinese AI startup DeepSeek has stirred the industry with its AI models, claiming they are on par with, or even outperform, industry-leading models from the U.S. at a fraction of the cost, reigniting the global AI race.
Baidu described its ERNIE X1 model as offering performance comparable to DeepSeek’s R1, but at half the price. The X1 is said to have enhanced capabilities in understanding, planning, reflection, and evolution, making it the first deep thinking model that can use tools autonomously.
Baidu also introduced its latest foundation model, ERNIE 4.5, which it claims offers “excellent multimodal understanding.” The model boasts advanced language abilities and improvements in understanding, generation, logic, and memory. It also has a “high EQ” and can easily understand network memes and satirical cartoons, according to Baidu.
Despite being one of China’s first tech giants to launch a ChatGPT-style chatbot, Baidu has faced challenges in achieving widespread adoption of its Ernie large language model, even though it claims performance comparable to OpenAI’s GPT-4, amid fierce competition.
Multimodal AI systems, like those developed by Baidu, can process and integrate various types of data, including text, video, images, and audio, enabling them to convert content across different formats.