-+ 0.00%
-+ 0.00%
-+ 0.00%

Excellent Ruixin (02687)'s mid-term revenue of 328 million yuan increased 19.1% year-on-year, and self-developed vertical AI model “Big Understanding” in the field of knowledge and technology completed the V5 version iteration

Zhitongcaijing·08/27/2026 13:25:04
Listen to the news

According to the Zhitong Finance App, Zhuoyue Ruixin (02687) announced interim results for the six months ended June 30, 2026. The group achieved revenue of 328 million yuan, an increase of 19.1%; gross profit of 165 million yuan, an increase of 28%; net loss of 116 million yuan, an increase of 17.6%; and a loss of 1.75 yuan per share.

During the reporting period, the company's self-developed vertical AI model “Big Understanding” in the field of knowledge technology completed the V5 version iteration, equipped with self-developed VOM (virtual object model) and VLM (visual language model), and connected to the World Labs world model to construct a full-modal capability matrix covering text, speech, vision, and 3D generation. At present, the service based on “Big Understanding” has covered a wide range of university users across the country, and consumes 800 billion tokens per month. Evaluation scores for core scenarios of knowledge application such as knowledge map construction, personalized learning path recommendations, knowledge reasoning question and answer, and search have steadily improved, indicating that the company has entered the large-scale application stage of AI capacity building.

During the reporting period, the company used a national education technology platform as the core anchor. Many achievements were recognized at the national level and formed leading demonstration effects, and the strategic card position continued to be deepened.

Facing medium- to long-term development, the company will seize the country's major industrial policy opportunities and continue to steadily advance the global business layout based on the three core barriers of structured intellectual assets across disciplines, self-developed full-modal models, and front-line in-depth delivery.