Baidu Unveils New AI Models to Compete in Global Tech Race
HONG KONG – Baidu, the Chinese technology giant, announced on Sunday the launch of two new artificial intelligence models, aiming to strengthen its position in a highly competitive AI landscape. Among these is a reasoning-focused model that Baidu claims rivals the innovative offerings from DeepSeek, a rising startup in the AI sector.
DeepSeek has recently made headlines by introducing AI models that they assert can compete with, or even surpass, leading models from the United States – all while being much cheaper. This development has stirred excitement and renewed competition in the global AI market.
Baidu highlighted one of its new models, ERNIE X1, stating that it delivers performance comparable to DeepSeek’s R1 but at only half the cost. The company notes that the X1 boasts improved capabilities in understanding, planning, reflection, and evolution, and it also represents the first deep-thinking model capable of using tools on its own.
Additionally, Baidu discussed its latest foundation model, ERNIE 4.5, which it describes as possessing excellent multimodal understanding. This model showcases advanced language skills and has significantly enhanced abilities in understanding, generating, logic, and memory.
Baidu emphasized that the new models also feature a high level of emotional intelligence, enabling them to comprehend internet memes and satirical images easily.
As one of the first major companies in China to roll out a ChatGPT-like chatbot, Baidu has found it challenging to achieve broad usage of its Ernie language model, despite claiming it performs similarly to OpenAI’s GPT-4 amidst stiff competition.
Multimodal AI systems, like the ones introduced by Baidu, are designed to process and combine different types of data, including text, images, video, and audio, allowing them to transform content across these forms seamlessly.
(Reporting by Farah Master and the Beijing newsroom)
