Kling Omni O3 on Genso AI: Multi-Segment Image-to-Video, End Frame & Optional Audio
Model guide for Kling Omni O3 (Kuaishou on WaveSpeed): 3–15s clips, optional sound, optional last frame, and chained segment prompts — how to set durations so segments add up.
What Kling Omni O3 is
Kling Omni O3 is Genso AI's flagship image-to-video line powered by Kuaishou's O3 stack on WaveSpeed. You supply a first-frame image (required path for typical runs), optional last-frame guidance, a global duration between 3 and 15 seconds, and a text prompt.
Differentiator: multi-segment prompts — you can chain up to eight segments, each with its own prompt and per-segment duration. The sum of segment seconds must equal the master duration you set for the clip. That pattern is ideal for story beats, shot-reverse-shot rhythm, or controlled beat changes inside one export.
Optional synchronized audio
Toggle Generate Sound when you want the model to produce synchronized audio alongside the video. Enabling sound uses a higher credit rate than silent renders — preview credits in the control bar before generating.
Use silent passes for rapid blocking; turn sound on when the dialogue or ambience is part of the approval criteria.
Aspect ratio and frames
Output aspect follows your first-frame image dimensions — pick your plate aspect before upload so downstream social crops stay predictable.
Last frame is optional; when used thoughtfully it gives the model a landing pose or composition for the final seconds.
When to pick O3 vs Kling 3.0
Pick O3 when segment-level storytelling, optional native audio, or tighter end-state framing wins. Pick Kling 3.0 when you want a simpler parameter sheet for hero character motion without segment math.
Both coexist in Video Lab — benchmark with the same first frame and prompt skeleton.
Ready to try it yourself? Free credits on sign up.
Try Kling Omni O3