A butterfly landing on lips, a bedroom turning into a nightclub, a mage calling up a phoenix. Each transition starts on one image and lands on the other, with both frames shown in the corners.
Start & end frame to video, also called first & last frame to video or keyframe video, turns two still images into one moving clip.
Start frame
End frameYou upload a start frame, the image the video opens on, and an end frame, the image it lands on. The AI model then generates every in-between frame, so the move from the first frame to the last frame plays as one smooth, continuous transition instead of a hard cut.
Because both keyframes are fixed, you decide exactly how the clip begins and ends. Regular image to video only locks the first frame and lets the model guess the ending. A start and end frame video generator is the better pick for before and after reveals, product reveals, scene morphs and transitions between two shots.
?Image to videoOnly the first frame is fixed. The model guesses the ending.
Start & end frameBoth ends are fixed. You decide how the clip lands.
Describe what happens in between: a slow camera push, a spin, a fold, a morph. The AI follows your prompt while keeping lighting and subjects consistent between the two frames.
Every start and end frame model generates sound together with the video, so a transition comes with matching ambience and effects instead of silence.
Every new account gets 15 free credits, no card needed. After that, the exact cost is shown on the Create button, and credits are only charged when a video is delivered.
Put your first frame in the Start slot and your last frame in the End slot. Matching framing gives the smoothest blend.
Write what happens between the two frames: the motion, the camera move and the mood. Then pick a model and length.
Click Create and the AI fills in every frame between your start frame and end frame. Download the MP4 when it's ready.
Every model below supports first & last frame to video. Start with the cheapest draft, then rerun the same pair on a stronger model.
| Model | Best for | Resolution | Length | From (credits) |
|---|---|---|---|---|
| Seedance 2.0 Mini | Cheap drafts | 480p–720p | 4–15 s | 15480p · 4s |
| Veo 3.1 Lite | Cinematic camera moves | 720p–4K | 4 / 6 / 8 s | 15720p · 8s |
| Gemini Omni Flash | Clean in-betweens | 360p–4K | 4 / 6 / 8 / 10 s | 451080p · 4s |
| Seedance 2.0 Fast | Faster Seedance renders | 480p–720p | 4–15 s | 47480p · 4s |
| Kling 3.0 | Character motion | 720p–4K | 3–12 s | 54720p · 3s |
| Seedance 2.0 | Realistic morphs | 480p–4K | 4–15 s | 76480p · 4s |
| Seedance 2.5 | Finest detail | 480p–1080p | 4–30 s | 112480p · 4s |
"From" is each model's lowest price. Most models charge more for higher resolutions or longer clips.
The cheapest way to check that two frames blend well, with sound. Fix the pair or the prompt here before paying for a bigger model.
Kling keeps people and their movement steady between the two frames, and the framing follows your images.
Rerun your best pair on Seedance 2.0 for smooth, realistic morphs, the most aspect ratios and up to 4K.
Room makeovers, hair and makeup, fitness progress or photo restoration. Put the before as the start frame and the after as the end frame.
Go from a closed box to the product in hand, or from a plain packshot to a styled scene, for ads and product pages.
Turn a sketch into a finished render, day into night, or summer into winter in one continuous shot.
Generate a bridge clip that starts on the last frame of one shot and ends on the first frame of the next.
Outfit swaps, glow-ups and "wait for it" reveals in 9:16 for TikTok, Reels and Shorts.
Start on an abstract frame and land on your logo or title card for a short branded opener.
Uploading frames and writing prompts is free. The cheapest start and end frame transition is Seedance 2.0 Mini at 480p: 15 credits for 4 seconds or 19 for 5. Veo 3.1 Lite is 15 credits per clip at 720p. 720p needs a Starter plan, and 1080p and 4K need Pro.
언제든 해지 가능 · 결제 기간이 끝날 때까지 이용
크레딧
기능
크레딧
기능
크레딧
기능
크레딧
기능
Upload your start and end frame, describe the motion, and let AI fill in everything between them.
Compare Plans
