Kuaishou Debuts Kling 3.0 to Lead AI Video Generation Race

The updated model, Kling 3.0, extends the maximum video duration to 15 seconds, a 50% increase over its predecessor. Beyond clip length, the upgrade focuses on complex scene composition and native audio generation. This allows AI-generated characters to speak naturally with varied accents and languages within a single clip, solving a technical hurdle that previously limited multi-character interactions.

Enhanced Realism and Audio Integration

According to a company statement released Thursday, the new architecture can manage three distinct characters speaking different languages in the same frame. This level of synchronization aims to close the gap between AI-generated content and traditional cinematography. The model is currently available to select subscribers, with a broader public rollout expected shortly.

Market Confidence and Global Competition

Financial analysts have responded positively to the update. Citi raised its target price for Kuaishou to HK$95.00 from HK$88.00, citing the "decent momentum" of the Kling platform. The bank expects the AI tool to drive significant revenue growth by 2026. Kuaishou’s shares recently closed at HK$73.00 in Hong Kong.

Kling remains a primary challenger to high-profile Western models, including Google’s Veo and OpenAI’s Sora. Within China, the company faces increasing pressure from startups such as MiniMax, which are also racing to commercialize high-fidelity video generation for the global market.