Chinese tech giant Baidu (BIDU) said it has launched two artificial intelligence (AI) models, including the ERNIE X1, which it claims delivers the same performance as DeepSeek R1 “at only half the price.”
Baidu said in a statement Sunday that it had released ERNIE 4.5—its native multimodal foundation model—and ERNIE X1, the “deep-thinking reasoning model with multimodal capabilities” that the company says rivals DeepSeek’s super-efficient open-source AI model. Both models, the Chinese company said, are free for individual users of its chatbot.
ERNIE X1, Baidu said, “possesses enhanced capabilities in understanding, planning, reflection, and evolution.” The deep-thinking reasoning model, Baidu said, excels in areas including dialogue, logical reasoning and complex calculations.
DeepSeek threw markets in disarray earlier this year when the Chinese startup released its own open-source AI model that performed as well as OpenAI’s ChatGPT and were built at a fraction of the cost using less advanced chips.
Baidu’s U.S.-listed shares are up around 1% in premarket trading Monday and have lost almost 10% of their value in the 12 months through Friday.