According to Forbes, PalFish has reached 70 million users globally and topped $100 million in revenue in 2024, with expectations to double that figure this year. The Beijing-based edtech company, founded by ByteDance co-founder Henry Huang, is expanding rapidly with 20% of revenue coming from international markets—primarily Southeast Asia—and that segment expected to grow 150% in 2025. PalFish International VP Xiao Li revealed plans to aggressively enter Middle Eastern markets within three years, building on their current 14 overseas centers in Thailand and Vietnam with five more planned in Southeast Asia by end of 2025. The company’s growth is powered by their AI tutor “Mia,” described as a “super teacher” capable of handling full-process classes including knowledge explanation, Q&A, personalized exercises, and emotional perception. Major investors include GGV Capital and Bertelsmann Asia Investment Fund.
The AI teaching revolution is real
Here’s the thing about PalFish’s approach—they’re not just building another chatbot. Their “Mia” AI tutor represents what they call “super teachers” that can independently run entire classes. That means from explaining concepts to generating personalized exercises based on real-time student feedback, and even detecting emotional states to provide encouragement. It’s basically trying to replicate everything a human teacher does, but at scale.
And scale is exactly what they’re achieving. When you can deploy AI teachers that work 24/7 across multiple languages and cultures, you fundamentally change the economics of education. Li claims this is “reshaping the cost structure of the education industry” where competition now depends on “understanding of technology, application speed, and product iteration capabilities” rather than capital scale. That’s a pretty bold claim, but with revenue doubling year-over-year, they’re certainly proving something’s working.
global-expansion-playbook”>The global expansion playbook
PalFish’s international strategy looks textbook smart. They’re starting in Southeast Asia where they’ve already established 14 physical centers, then targeting the Middle East—both regions with growing middle classes willing to invest in education. What’s interesting is how they’re using AI not just for teaching but for managing their own global operations. They mention using AI to “optimize cross-cultural team management” and achieve “precise collaboration among global teams.”
Now, that’s the kind of operational efficiency that could give them a real edge as they scale. When you’re dealing with teams across different time zones and cultures, having systems that can streamline communication and decision-making becomes crucial. It’s one thing to have great technology for your customers—it’s another to use technology to run your own business better.
Moving beyond English
PalFish is already expanding from English into math and other subjects, which makes complete sense. Once you’ve built the AI teaching infrastructure, adding new subjects becomes more about content than technology. Their course pricing—ranging from $200 to $5,000 in different Southeast Asian markets—shows they’re targeting various income segments.
But here’s what I find most compelling: Li’s argument that we’re creating “AI natives”—children who learn to collaborate and create with AI from an early age. That’s a fundamentally different vision than just using technology to deliver traditional education more efficiently. It suggests we’re looking at a generation that will interact with AI as naturally as previous generations interacted with computers or the internet.
The human element still matters
Despite all the AI talk, PalFish still maintains “care teams” that provide feedback to parents and learners based on AI-generated reports. And Li himself uses the products with his family, getting that crucial real-world feedback. That combination of high-tech AI with human oversight seems like the sweet spot—technology handles the scalable parts while humans handle the nuanced, emotional intelligence aspects.
So where does this leave traditional education companies? If PalFish’s numbers are accurate, we’re looking at a company that’s figured out how to scale quality education globally in a way that was previously impossible. The question isn’t whether AI will transform education—it’s how quickly the incumbents can adapt before companies like PalFish eat their lunch.
