When AI Picks Your College: How Baidu, Alibaba, and Tencent Are Battling Over 12.9 Million Students

On June 10, 2026, just hours after China's national college entrance exam ended, three tech giants launched their AI college advisory tools almost simultaneously. The battlefield was not search ads or e-commerce — it was 12.9 million exam takers and their families.

This is more than a story about education. It is a story about AI, platforms, and the future of digital services — and it has implications for anyone advertising on Baidu.

🎮 The Three Players

All three are free for students. All three use large language models. But their approaches — and what they signal about each company's AI strategy — are very different.

🤖 Baidu's Play: AI + Human Trust

Baidu's AI College Report pulls data from years of admission scores, 2,000+ university programs, and employment databases. The AI generates a personalized report with recommended schools, risk analysis, and a clear explanation of the logic behind every recommendation.

But Baidu added a twist: every AI-generated report is reviewed and certified by a human admissions counselor.

The same model appears in Baidu's Merchant Agent (商家智能体), where AI handles initial customer conversations and seamlessly hands off to human agents. The pattern is consistent: Baidu positions AI as augmentation, not replacement.

📱 Alibaba's Play: Pure AI Agent

Alibaba's Qwen Exam Agent takes a different approach. Built on Qwen's large model and Quark's 8 years of college application data, it is designed to be fully autonomous:

• Students input their estimated scores and subject preferences
• The AI generates a customized report with dozens of school combinations
• Students can interact in real time, adjusting preferences and receiving updated recommendations

In 2025, Quark's AI model already provided 13 million free admission reports. This year, Alibaba upgraded to a full Agent architecture, claiming it can replace services that typically cost over ¥5,000 in the market.

💬 Tencent's Play: Conversational Discovery

Tencent's Yuanbao Exam Guide, powered by the Hunyuan large model, takes a conversational approach. Rather than generating a fixed report, it engages students in multi-turn dialogue:

• "What kind of work environment do you prefer?"
• "How important is the school's reputation vs. the program's ranking?"
• "Are you open to studying in a less popular but emerging field?"

The AI dynamically adjusts school recommendations and re-balances the "reach / match / safety" ratio as the conversation evolves. Tencent's philosophy: the best AI advice comes from understanding the individual — and that requires conversation, not just computation.

🔍 What This Battle Reveals About Baidu

For advertisers, this story matters because it reveals how deeply AI is penetrating every corner of Baidu's ecosystem.

Baidu's AI College Report is built on the same foundation as its ad products:

• Wenxin Assistant (文心助手): The conversational AI that powers college advisory is the same technology that powers Baidu's search and recommendation algorithms
• Human-AI collaboration: The "human review" model in college advisory mirrors the Merchant Agent's human handoff feature
• Data ecosystem: The employment database, admission scores, and user behavior data that fuel the College Report are part of the same data infrastructure that powers Baidu's ad targeting

📋 The Real Competition: AI Super-App

What is happening here is bigger than college admissions. Baidu, Alibaba, and Tencent are racing to build AI super-apps — platforms that can handle everything from search and shopping to education and healthcare through a unified AI interface.

The college exam advisory battle is a microcosm of this larger war. Each company is testing a different AI strategy in the same real-world laboratory: 12.9 million students making one of the most important decisions of their lives.