From "Ten Thousand Yuan Consultation" to Quark Free AI: A Battle for "Information Equality" in College Entrance Examination Volunteering

Author: Zhang Yongyi

On June 10, the last exam of the 2025 college entrance examination has concluded. However, for millions of families with candidates across the country, the upcoming college application process is another kind of college entrance examination.

Faced with the admission information from over 3,000 universities and visions for future life, candidates and parents are becoming more concrete in their considerations of reality. The market has also begun to mass-produce different types of volunteer application products and services to meet the personalized needs of those born after 2005.

From the planning services that cost tens of thousands of yuan offered by "Zhang Xuefeng" and the "low-config" tutoring priced at several thousand yuan on social platforms, to the "AI software" available for hundreds of yuan, the commercialization wave of college entrance examination application is gradually reaching its peak.

Many AI products are starting to dismantle the information barriers in front of candidates by leveraging their capabilities in data and large models. Large models have appeared in the college application scenario, allowing more candidates and parents to achieve information equality.

As a platform that has been deeply engaged in Gaokao information services for seven consecutive years, Quark once again stands behind the candidates this year—not only releasing the industry's first Gaokao volunteer large model and Gaokao knowledge base, but also launching AI-centric features such as "Volunteer Report" and "Gaokao Deep Search."

Quark's goal is clear: to effectively address every open-ended question related to college application through product innovation and AI technology. At the same time, provide each candidate with a professional application report to assist them in making life decisions.

01 How to create a good "personalized" volunteer report?

In real-life volunteer application services like those provided by Zhang Xuefeng, teachers must first master a large amount of data and exclusive information to build their own moat. Then, through one-on-one questioning, they gain a deep understanding of the candidates' information, interests, and family situation, allowing for analysis and judgment. After modifications and comparisons, a "Volunteer Report" will be delivered as the final result to the candidates who are the users.

What will an AI-generated "volunteer report" look like in the face of such a complex scenario with extremely high information density and a very long decision-making chain?

After Quark launched the "Volunteer Report," I personally experienced this feature. Taking a candidate from Beijing with a score of 630 in Physics, Chemistry, and Biology as an example, we set his interest in law and his desire to become a lawyer. First, we fill in his personal information and preferences, and complete the personal profile by answering 12 questions.

After clicking confirm, Quark will start generating the report, which will take about 5-10 minutes and consist of 15-20 pages.

In this process, Quark will rely on a large model for college entrance examination preferences to provide personalized planning advice for candidates through Agent calls. Finally, three different specialized reports will be output (major priority, institution priority, and regional priority), which include content such as strategy design, detailed information on institutions and majors, and interpretation of the preference list. Users can directly add them to the preference list or export them as PDF.

From the results of the report, it is able to understand my preferences in law and develop a gradient plan in conjunction with 985, 211 universities and specialized disciplines.

It can also analyze the situation of cities, tuition fees, and professional employment, integrating them into the recommended schools and majors, allowing candidates to grasp all information more clearly, rather than just simple university and major information.

Data shows that only 2% of candidates choose to seek offline consultations each year. For the 98% of candidates, the emergence of Quark has alleviated concerns about geography and costs, reducing the information gap in filling out applications.

In addition, with the support of the deep search function for the college entrance examination, even some open-ended and highly colloquial descriptions can provide more realistic reference suggestions.

For example, the author used "Shandong Physics and Chemical Boy, 647 points, recommend 985 schools that you can go to, you can run schools with Sino-foreign cooperation, and if you want to go to graduate school or study abroad in the future, do a voluntary fill-in" as a prompt word, and experienced the in-depth search ability of the college entrance examination.

Under this prompt, Quark will first analyze the core demands of the examinee—scores, subject selection, interests, and regional preferences—then perform multidimensional matching and reasoning within its vast college entrance examination knowledge base. This knowledge base not only contains structured data such as historical admission data of institutions, major information, employment rates, and further education rates, but also incorporates a large amount of unstructured knowledge about industry development trends and the correlation between majors and professions.

How is the "Expert Brain" cultivated?

In order to make the user experience of these front-end users more accurate, Quark has chosen to further invest in the expansion of AI model capabilities this year, with a 100-fold increase in computing power investment.

Although it is based on the Universal Q&A model, the Quark College Entrance Examination Volunteer Model is not simply a fine-tuning of the general model. Instead, it employs a strategy refinement mechanism guided by value judgments made by real college entrance examination volunteer experts, allowing the model to truly "think like a volunteer expert and provide suggestions."

To achieve this goal, Quark first needs to teach AI to imitate the "thinking chain" of real human experts. During the instruction fine-tuning phase, the R&D team structured the multi-turn real conversations between hundreds of experienced college entrance examination counselors and students, and parents, extracting complete analysis pathways and communication language styles. These high-quality supervised data, which include tens of thousands of real expert "reasoning chains", serve as the "textbook" for the large model to learn the analysis processes of human experts.

In this regard, Quark has demonstrated its core advantages. "The data used by Quark comes from authoritative materials released by the official examination institutes, similar to the industry-recognized 'Big Thick Book,'" emphasized Mr. Ren, an expert involved in training the Quark AI volunteer model. This stands in stark contrast to many large models that rely on unverified old data scraped from the internet, fundamentally eliminating absurd AI illusions such as 'a student with a score of 500 being recommended to a 985 university,' and ensuring the accuracy and authority of the recommendations.

03 Pulse of the Era

From "How the AI College Entrance Examination Volunteer Model is Formed", you must have noticed that such a large model rooted in specific application scenarios can produce accurate results for students, thanks to the assistance of real college entrance examination volunteer experts.

Every year during the college entrance examination season, we see more professional volunteer guidance experts like Zhang Xuefeng. However, in reality, there are still many "volunteer filling mentors" of questionable quality targeting anxious students and parents.

The reason why these services continue to thrive every year is that the significance of voluntary enrollment has far surpassed "choosing a school and a major"; it has evolved into the "first career planning" in which the entire family participates. It can even be said to have become the "pulse of the times" shared by tens of millions of candidates every year.

However, what Zhang Xuefeng represents is an expensive solution at the top of the pyramid. A single Zhang Xuefeng has limited energy, and his services are destined to be a "luxury" for a small number of people. Behind him is a much larger market of mixed quality, where countless institutions and individuals claiming to be "experts" offer services of questionable quality, reaping the benefits from ordinary families who are equally anxious but cannot access top resources.

If Zhang Xuefeng's core barrier is his personal experience and information accumulation, then Quark's approach is to internalize the decision-making logic and experience of hundreds of senior planners through a large model of college entrance examination choices, combined with the largest and real-time updated knowledge base for college entrance examinations in the country, attempting to transform the past reliance on personal, expensive, and non-standard "expert services" into a standardized and high-quality "AI advisor" that every ordinary person can access for free.

Some people say that the AI college application tool launched by Quark this time is "going to overturn the table", but that's not quite accurate. It’s not about overturning the table—it’s about replacing it with a bigger table, allowing more people to sit at it. This table doesn’t require reservations and doesn’t charge consultation fees. As long as you can open your phone and fill out a profile, it can provide you with a truly logical and data-driven application report.

On this table, the candidates from Liangshan and Hangzhou see the same report structure, the same professional dimensions, and the same recommendation logic. Their starting point has been brought a little closer by AI. The college entrance examination is an opportunity to change one's destiny, and the significance of technology is to make "opportunity" a little fairer. Such opportunities should never be placed solely on a VIP gold card table.

The user data publicly disclosed by Quark at the press conference on June 12 provides the best commentary on this - its college entrance examination service has helped a total of 120 million users, with users from third-tier cities and below accounting for more than 50%.