Step 1
Upload the first frame
Add the image where the shot should begin. Sign in first; JPG, PNG and WebP files are supported.
First and Last Frame to Video Generator
Upload where the shot starts and where it ends. AI animates everything in between.
Genreel / Director deskWan 2.2 Fast · 480P · 5s
Direct the frame. Choose the engine.
30+ models in one workspaceCost shown before you generateFailed renders refunded
Examples
What Is First and Last Frame to Video?
First and last frame to video is a form of image-to-video generation where you supply two images instead of one. The first image becomes the opening frame, the second becomes the closing frame, and the AI model generates the motion that connects them.
Because both ends are fixed, you decide exactly where the shot lands. That makes it far more predictable than a single starting image, where the model chooses how the clip ends.
On Genreel it works with Seedance 2.5, Kling 3.0, Veo 3.1, Veo 3.1 Fast, Wan 2.2 Fast and PixVerse V6. If you only have one image, the standard image to video generator is the better fit.
How to Use First and Last Frame
Two images and one sentence of motion are all it takes.
Step 2
Add the last frame
Upload the closing image. Keep the same aspect ratio and a similar subject position so the transition stays smooth.
Step 3
Describe the motion
Write how the first frame turns into the last — "the box opens and the sneaker rises out", "the camera pulls back as day turns to night".
Step 4
Pick a frame-capable model and generate
Choose a model that supports a last frame, set the duration, and generate. Download the MP4 when it is ready.
When to Use It
Fixing both ends of a shot is most useful when the destination matters as much as the motion.
Product reveals and morphs
Start on packaging and end on the product, or move from a sketch to the finished design. The final frame is exactly the shot you approved.
Scene transitions
Link two locations, two seasons or day and night with one continuous move, so a sequence of clips cuts together cleanly.
Animation in-betweening
Give two key poses of a character and let the model draw the motion between them — a quick way to test animation timing.
Planned camera moves
Use a wide shot and a close-up of the same scene to get a push-in that lands exactly on your subject, instead of a camera that drifts wherever the model wants.
Seamless loops
Upload the same image as first and last frame and ask for gentle motion in between. The clip returns to where it started, so it loops cleanly on a website or social post.
For the cleanest results, keep the lighting and framing of both frames close. Large jumps in angle or style give the model more to invent.
First and Last Frame FAQ
Quick answers about generating video from two keyframes.
Can I use only a first frame?
Yes. The last frame is optional — with only a first frame, the model decides how the shot ends. That is standard image to video. A last frame always needs a first frame.
How long can the video be?
It depends on the model: Seedance 2.5 offers 4 to 30 seconds, Kling 3.0 3 to 15, Veo 3.1 4, 6 or 8, and Wan 2.2 Fast 5 seconds. Short clips give the smoothest transitions.
Will the middle of the video stay consistent?
Usually, when both frames share the same subject, lighting and aspect ratio. If the two images differ a lot, describe the change in the prompt so the model knows what to transform.
How much does it cost?
The same as a regular generation on the chosen model. Credits depend on model, length and resolution, and the price is shown before you generate. New accounts start with 40 free credits.
Which models support first and last frame?
Seedance 2.5, Kling 3.0, Veo 3.1, Veo 3.1 Fast, Wan 2.2 Fast and PixVerse V6. The upload option for the last frame only appears when the selected model supports it.
Set the start. Set the end.
Upload two frames and let AI animate the path between them.