Zheng Quan Shi Bao Wang·2025-12-04 03:14Core Insights - Kuaishou's Keling AI has launched the video generation model 2.6, which introduces the "audio-visual synchronization" capability, enhancing the traditional AI video generation workflow [1] - The model upgrade includes two main functions: text-to-audio-visual and image-to-audio-visual, significantly improving creative efficiency [1] - Keling AI aims to achieve $140 million in revenue by 2025, exceeding the initial target of $60 million by over 100% [2] Group 1 - The Keling AI model 2.6 can generate complete videos with natural language, sound effects, and ambient sounds in a single output, marking a significant advancement in AI video creation [1] - The model demonstrates strong performance in audio-visual coordination, audio quality, and semantic understanding, maintaining a global lead in Chinese voice generation [1] - Kuaishou's CEO emphasizes the vision of enabling everyone to tell great stories using AI, focusing on AI film creation as a core objective [2] Group 2 - Kuaishou is increasing its capital expenditure (Capex) to enhance computing power for Keling AI, responding to growing user demand for video generation models [3] - The company plans to achieve a mid-to-high double-digit growth in overall Capex spending by 2025 compared to the previous year [3] - Kuaishou is also investing in other AI business capabilities alongside Keling, indicating a broader commitment to AI technology integration [3]