According to Smart Emergence, Meituan, the leading food delivery company, has taken an important step in its AI strategy. Pan Xin, the former head of the visual model AI platform at ByteDance, has officially joined the company to lead innovation in multimodal AI. This personnel change has drawn widespread attention in the industry, marking Meituan's efforts to enhance its AI technology infrastructure amid increasingly fierce competition in the food delivery market.

With rich experience in the field of artificial intelligence, Pan Xin previously worked as a researcher at Google Brain and participated in the development of TensorFlow's dynamic graph mode. After returning to China, he held key positions at tech giants such as Baidu, Tencent, and ByteDance, where he was responsible for building AI platforms in the multimodal field, accumulating deep technical expertise. After joining Meituan, he will lead the development of multiple applications related to multimodal AI, including the LongCat App.

Meituan

In Meituan's 2025 strategy, AI is viewed as a key factor in enhancing core competitiveness. Founder Wang Xing stated that Meituan's strategic positioning on AI technology is proactive offense rather than passive defense. Meituan's AI strategy mainly includes three aspects: applying AI to employee work and operations, using AI to improve existing products and services, and heavily investing in the development of self-researched large language models (LLMs).

However, since the second quarter of 2025, the competition in the food delivery market has intensified, and Meituan has invested more resources in training AI foundation models, hiring a large number of AI talents from big companies like Alibaba and Tencent. Despite external pressures, Meituan's enthusiasm for AI technology research and development has not diminished, and its recruitment standards remain high.

Since the beginning of 2025, Meituan has successively launched multiple AI foundation models and applications, including the AI coding product Nocode, the large language model LongCat-Flash-Chat, and the video generation model LongCat-Video. The release of these products demonstrates Meituan's comprehensive layout in AI technology, expanding from language models to full modalities such as vision, audio, and video.