(Bloomberg) — Meituan unveiled plans for the first time to develop its own AI model after reporting a 20% jump in quarterly revenue, becoming the latest big Chinese tech player to join an accelerating global wave of development.

Most Read from Bloomberg

China’s meal delivery leader is starting to integrate AI into work processes and across its range of consumer services, which include a travel site and e-commerce platform. It’s developing its own “LongCat” large language model, a platform that will compete with rivals such as ByteDance Ltd.’s Doubao and Alibaba Group Holding Ltd.’s Qwen.

Meituan is a surprise entry in a quickening race that spans OpenAI and Meta Platforms Inc. in the US to Baidu Inc. and DeepSeek in China. While the company has long relied on deep data analysis and algorithms to match merchants and consumers, it’s said little about its AI strategy. Meituan now intends to ramp up capital spending on AI to take on rivals directly, founder Wang Xing said. The company is investing billions in AI chips, Wang added without specifying amounts.

LongCat already performs well against domestic industry leaders and should form the basis of a broad-based AI strategy that includes bridging the online and physical markets, such as in food delivery, the billionaire founder said.

“Our strategy is offense, not defense,” Wang told analysts on a conference call. “When something as fundamentally revolutionary as AI is coming, the only strategy that makes sense is not trying to defend the way you already have.”

Wang’s pronouncement came after Meituan posted a 20% jump in revenue to sales of 88.5 billion yuan ($12.2 billion) in the December quarter, about 1% ahead of the average projection. Net income surged to 6.2 billion yuan. That suggested the Chinese meal delivery leader is successfully fending off new domestic competition while expanding abroad.

That’s also giving Meituan ammunition to stake out a place in the AI arena. The company benefits by having one of the more extensive real-world networks of people, merchants and users among its peers, Wang argued.

“To bring AI to the physical world you need more than just very smart app algorithms or models. You need infrastructure in the physical world and that’s our advantage,” Wang said.