每天玩近2小时，县城青少年被AI角色拿捏了

Core Insights - The article discusses the innovative approach of the AI character brand "Zao Meng Ci Yuan" (造梦次元), which has rapidly gained traction in the AI content community by transforming model capabilities into consumable content experiences [5][7][54] - The company has achieved significant user engagement, with a daily average token consumption reaching 200 billion at peak times and over 10 million cumulative users by mid-2025, primarily consisting of young females aged 12 to 35 [7][13][21] Group 1: Company Overview - "Zao Meng Ci Yuan" is positioned as an AI content community that focuses on converting model capabilities into engaging content experiences [7][54] - The company started with AI interactive storytelling, allowing users to create narratives with AI assistance, and has since expanded to include various content forms such as music and video [11][12] - The founder, Shen Qiajin, emphasizes the importance of character IP as the core value of light interactive content, likening their IPs to internet celebrities [14][62] Group 2: User Engagement and Demographics - By June 2025, the platform had over 10 million users, with a daily average engagement time of 1 hour and 50 minutes, indicating high user retention [13][54] - The primary user demographic includes young females from lower-tier cities, with a significant overlap with users of platforms like Douyin and Kuaishou [21][32] - The company aims to attract a broader user base by evolving its content offerings in line with advancements in AI model capabilities [39][52] Group 3: Product Development and Strategy - The product design focuses on creating a versatile platform that can adapt to various content forms without being limited by specific user demographics or AI characteristics [66][68] - Shen Qiajin stresses the need to align product development with real user needs and feedback, avoiding the trap of catering to overly critical users who do not represent the majority [27][28][76] - The company is actively working on integrating multi-modal content to attract a wider audience, particularly older users who prefer video content [52][53]