Seedance 2.0 First to Last Frame: Video Between Two Images
Last verified: November 3, 2026
Seedance 2.0 generates sound and picture together through a joint audio-video architecture, so the shot it builds between two of your stills can come back with its soundtrack already in place. With Seedance 2.0 first-to-last-frame generation, you set the opening image in First Frame and the closing image in Last Frame, describe what happens in between in the Prompt box, keep Generate Audio on, and render at up to 4K when the clip is going into a finished edit.
Your two images fix where the shot starts and where it ends, so the prompt only has to cover what happens in between: what moves, how the camera travels, how the light shifts and what the audience hears. That split suits Seedance 2.0, which is built for substantially improved controllability in video generation. Write the prompt as a director's note about the change, not a description of pictures you have already supplied.
Explore more first to last frame
Seedance 2.0 Keyframe Video Settings at a Glance
Inputs, clip lengths, resolutions, frame shapes and output for each run.
Input
First Frame (JPG, JPEG, PNG or WEBP, up to 10 MB) + Last Frame (JPG, JPEG, PNG or WEBP, up to 10 MB) + a text prompt of up to 2,048 characters
Video Duration
4 to 15 seconds
Resolution
480P, 720P, 1080P or 4K
Aspect Ratio
1:1, 3:4, 4:3, 9:16, 16:9
Audio
Switch on or off, on by default
Output
1 video per run
Techniques for Better Seedance 2.0 Keyframe Clips
Sound Cues Written Into the Prompt, Tied to Visible Action
One Named Camera Move From the First Framing to the Last
How to Turn Two Images Into One Seedance 2.0 Clip
Six steps take you from two uploaded images to a downloaded video.
Upload the opening image in "First Frame"
Click "Upload first frame" and choose the image the clip should open on. Treat it as the first thing viewers see: the subject, setting and light before anything changes, framed the way you want the shot to begin.
Upload the closing image in "Last Frame"
Click "Upload last frame" and add the image the clip should end on. Keep the subject and setting recognisably the same as in your first image, and change only what the shot is about, such as the light, a pose or the camera position. That leaves your prompt one clear change to describe instead of several competing ones.
Describe the movement in the "Prompt" box
Write what happens between the two images (the action, the camera move and the sound) in up to 2,048 characters. Don't re-describe what the images already show; spend the space on the change itself. If you're starting from a rough idea, the "Enhance Prompt" button under the box suggests fuller versions, and your text only changes when you pick one.
Set "Video Duration", "Resolution" and "Aspect Ratio"
"Video Duration" runs from 4 to 15 seconds and starts at 4. A quick light change or gesture fits the short end, while a camera move combined with action needs more room. "Resolution" offers 480P, 720P, 1080P and 4K and starts at 480P, which is enough to judge whether the motion works; choose 1080P or 4K for the take you'll publish. "Aspect Ratio" offers 1:1, 3:4, 4:3, 9:16 and 16:9 with 1:1 preselected, so pick the shape of the platform you're posting to and prepare both images in that shape.
Check "Generate Audio" and "Web Search", then click "Generate"
"Generate Audio" is on by default and gives the clip a soundtrack; switch it off when the clip will get its music or voice-over in your edit. "Web Search" is off by default and turns on online search for the run; leave it off unless your prompt mentions something you want looked up. Then click "Generate".
Download the result
Each run returns one video. Before you download it, check three things: that it opens on your first image, that it ends on your last, and that the movement in between holds together. If the middle drifts, refine the prompt or give the clip more seconds and generate again.
Two Keyframes or a Text Prompt: Choosing a Seedance 2.0 Workflow
How starting from two images compares with describing the whole shot in words.
| Criterion | Seedance 2.0 | Alternatives | Best For |
|---|---|---|---|
| What you need before you start | You need two finished images, one showing how the shot opens and one showing how it ends, plus a prompt for the movement between them. | Seedance 2.0's text-to-video workflow starts from a written prompt, so there are no images to prepare before the first run. | Pick the two-image workflow when your opening and closing pictures already exist, and text to video when you're still working out what the shot should look like. |
| How the ending is decided | The image you upload in Last Frame sets where the clip ends, so you choose the closing composition before you generate. | In text to video, the closing composition comes from how Seedance 2.0 reads your description, so you set the ending in words instead of with a picture. | Choose the two-image workflow when a shot has to end on a specific picture, such as a product pack shot or a storyboard panel, and text to video when the ending can stay open. |
Common Problems With Seedance 2.0 Keyframe Clips and How to Fix Them
The issues that come up most often when turning two images into one clip.
My image won't upload to First Frame or Last Frame.
My video was blocked after I uploaded a photo of a real person or a famous character.
My video came out square even though both images are widescreen or vertical.
The change between my two images happens too fast and the motion looks rushed.
Frequently Asked Questions
Do I need both a first frame and a last frame?
Yes. First Frame and Last Frame are both required, because the generator builds the clip between the two images you supply. If you don't have images yet and want to describe the whole shot in words, switch to Seedance 2.0 Text To Video.
How long can a first-to-last-frame video be?
Up to the longest single clip Seedance 2.0 is documented to generate. Video Duration covers the model's full documented range from its shortest clip to its longest, so you don't give up any length by working from two frames; the exact options are listed in the Video Duration step of the how-to. If a change needs more time than one clip holds, the troubleshooting section shows how to split it across two runs.
Can I make vertical videos for TikTok, Reels or Shorts?
Yes. Aspect Ratio includes a vertical option alongside square, widescreen and two in-between shapes. Prepare both images in portrait orientation and pick the vertical setting before you generate, so the clip comes out in the same shape as the pictures it starts and ends on.
What image files can I upload?
Each run takes two images, one in First Frame and one in Last Frame, and both are required. Each slot accepts JPG, JPEG, PNG or WEBP files up to 10 MB, and each image must be at least 300×300 pixels.
What makes one generation cost more than another?
The credit cost of a run is calculated from the options you choose and shown above the Generate button once you're signed in. It depends on Video Duration and Resolution, so changing either of them changes the cost you see before you generate.