Which AI video models let you set both the first and last frame?
First-and-last-frame control means supplying two images and letting the model generate the motion between them, which pins where a shot ends instead of leaving it to chance. On Floniks the models supporting it include Kling in pro mode, Hailuo 2.0, Wan 2.7 with two input images, LTX 2.3, MiniMax H3 and Seedance 2.0 in keyframe mode. Models given only one image treat it as the opening frame.
Why the end frame matters
Given one image and a prompt, a video model decides for itself where the motion goes. That is fine for a standalone clip and a problem for a sequence, because the next shot has to start somewhere sensible. Pinning the last frame turns generation from something you sample repeatedly until it lands into something that reliably ends where the edit needs it. For anything cut together from multiple clips, this single feature saves more re-renders than any prompt technique.
How to use it well
The two frames should be plausibly connected by the motion you are asking for. A first frame of a closed door and a last frame of a beach will not produce a coherent transition; a closed door and an open door will. Generate both frames with an image model first, keeping the subject consistent between them by conditioning both on the same references, then hand the pair to the video model.
Chaining it into a sequence
The natural extension is to reuse each clip’s last frame as the next clip’s first frame, which produces a continuous sequence rather than a set of disconnected shots. In a workflow graph this can be wired once and rerun for every scene, so a multi-shot sequence becomes one execution rather than a manual relay of exports and uploads.
Related questions
Build it on Floniks
Image, video, digital humans, and reusable workflows on one canvas. No card required.
Explore Floniks