China's Baidu Unveils Two New AI Models

Baidu, a Chinese tech company, has announced the launch of two new artificial intelligence models, including a reasoning-focused model designed to rival DeepSeek's technology as reported by Reuters.
DeepSeek, a rising Chinese AI startup, has gained attention by developing models that it claims are comparable to top-tier AI systems in the United States while being significantly more cost-effective. According to Reuters, Baidu, in response, stated that its new reasoning model, ERNIE X1, matches the performance of DeepSeek R1 but at half the cost.
The company highlighted the X1 model’s advanced understanding, planning, reflection, and evolution capabilities, positioning it as the first deep-thinking AI model capable of using tools autonomously.
Improved Multimodal AI Capabilities
In addition to ERNIE X1, Baidu launched ERNIE 4.5, its newest foundation model with better multimodal comprehension. The firm highlighted that the model has improved language functions and better comprehension, generation, logic, and memory functions. ERNIE 4.5 is also created with a "high EQ," which allows it to better understand internet memes and satirical cartoons.
Challenges in the AI Market
Reuters reports that despite being one of China's first companies to launch a ChatGPT-style chatbot, Baidu has faced difficulties in achieving widespread adoption for its ERNIE large language model. While the company asserts that its AI technology is on par with OpenAI’s GPT-4, it continues to face fierce competition in the rapidly evolving AI sector.
The Rise of Multimodal AI
Multimodal AI models such as ERNIE 4.5 are capable of processing and combining various forms of data, for example, text, video, images, and audio, to enable fluid content conversion between the media. This milestone speaks to the growing emphasis on designing versatile AI solutions that are capable of interacting through various media.