Published on December 15, 2025
We are excited to announce the immediate availability of the Kling o1 Start/End Frame tool on RunDiffusion. This powerful new model brings a new level of control to AI video generation, prioritizing high fidelity and exceptional consistency by allowing creators to define the exact visual boundaries of their clips.
Anchor Your Narrative
Randomness has its place, but professional storytelling requires precision. With Kling o1, you must provide a start frame image to anchor your visual sequence. This ensures the video begins exactly as you envisioned, maintaining character and environmental consistency from the very first frame.
Bridge the Gap
For ultimate control over the visual arc, you can also provide an optional end frame image. The model then works to bridge the gap seamlessly, animating the transition between your starting point and your desired conclusion over a 5-second duration. This is ideal for:
- Creating specific object transformations.
- Moving a camera from point A to point B.
- Transitioning between two distinct lighting states.
Integrated Prompting Workflow
Kling o1 allows for deep integration of visual inputs into your textual description. You can reference your uploaded images directly in your prompt by typing @image1 for the start frame and @image2 for the end frame, giving the model clear context on how to handle the transition.
