Kuaishou’s Kling AI has opened the developer API for Kling 4.0, the flagship video generation model that launched globally earlier this month. The API service officially launched on October 22, 2026, and is being rolled out gradually to enterprise customers and developers worldwide.
Kling 4.0’s headline upgrade is native 30-second generation in a single pass, doubling the ceiling of the previous 3.0 model. It also adds up to 10 keyframes for shot-level control, Omni Reference with up to 15 combined reference inputs across images, video, voice and subjects, and native two-channel stereo audio with tighter lip-sync. Kling says the new model delivers more stable motion along with complex, director-style camera moves such as push-ins, tracking shots and orbits.
Among the first platforms to integrate the new API is SelfyzAI, an AI video and image generation platform that says it took part in Kling 4.0’s first-batch internal beta review ahead of release. The model is now available in the SelfyzAI workspace model picker alongside the platform’s existing lineup, including Seedance 2.5.
AI video generation is moving fast: Google’s Veo, OpenAI’s Sora and Genmo are all racing to ship longer, more controllable takes. For developers, a public Kling 4.0 API means these capabilities can now be embedded directly into production apps and creative workflows.
Source: EIN Presswire; GlobeNewswire via WCIA
