Skip to main content
NEW HERE30% OFFyour first month on any monthly plan, for 72 hours after you sign up
HomeAI VideoFirst & Last Frame to Video
Model12 models

First & Last Frame to Video AI

Add where it starts and where it ends. The model fills in the 5 to 8 seconds between, so the last shot is exactly the one you chose.

ExampleThe photo that went in

Seedance 2 · 5sShe spins once and her outfit and hair change mid-turn, smooth transition, camera stays still.

Start from a recipe

Each card keeps both frames, the words and the model. One tap loads the start frame, the end frame and the prompt.

TransformationA Latina woman in her twenties with a messy bun in an oversized grey hoodie and leggings posing in front of a tall mirror in her sunlit bedroom, full body, vertical.1The same woman in the same pose in front of the same mirror, now in a black satin evening dress with her hair down in loose curls and gold earrings; same mirror, same light.2

Spin into a new outfit

She spins once and her outfit and hair change mid-turn camera stays still smooth transition 5 s

Seedance 2 720P 5s 200
MakeoverAn empty rental living room with white walls, oak floor and a large window, afternoon light, no people.1The same living room fully furnished: a linen sofa, a rug, plants, a coffee table and framed prints; same window, same light and angle.2

Empty room to styled

Furniture appears piece by piece until the room is styled camera stays still smooth time-lapse 5 s

Seedance 2 Mini 480P 5s 30
Time-lapseThe front of a small brick house with a green door in the daytime, a bare tree and a short path, no people.1The same house at night covered in warm Christmas lights with a wreath on the green door and snow on the path; same angle.2

Porch, day to night

Day turns to night camera stays still snow falls and the lights switch on one string at a time 5 s

Seedance 2 Mini 480P 5s 30
CoupleA couple in their thirties, a Vietnamese man with short black hair and a white woman with a blonde bob, in jeans and T-shirts standing on church steps.1The same couple in the same pose on the same church steps, now in a tuxedo and a lace wedding dress, holding a bouquet.2

One step into the wedding

They take one step down and their clothes change into wedding outfits camera stays still petals fall 5 s

Seedance 2 Mini 480P 5s 30

How first and last frame works

Two pictures fix the ends. Your words describe the middle.

A couple in their thirties, a Vietnamese man with short black hair and a white woman with a blonde bob, in jeans and T-shirts standing on church steps.The same couple in the same pose on the same church steps, now in a tuxedo and a lace wedding dress, holding a bouquet.

Add the start and the end

JPG, PNG or WEBP · up to 10 MB. Two photos of the same scene from the same spot. The first one sets the shape of the clip.

They step down, clothes turn into wedding outfitswhat changescamerawhat else moves

Say what happens between

Name the change and keep the camera still. The model already knows both ends.

Get the in-between

Ready in one to four minutes, landing on your last frame.

Start, middle, end

The two frames on the outside went in. The clip in the middle is what the model made to join them.

The start frame
Start · Bare walls, one window
Seedance 2 Mini · 480P 5s fills the middle
The end frame
End · Sofa, rug, plants, art
The start frame
Start · Grey afternoon
Seedance 2 Mini · 480P 5s fills the middle
The end frame
End · Snow and every light on
The same couple in the same pose on the same church steps, now in a tuxedo and a lace wedding dress, holding a bouquet.1One step down2Wedding clothes3Petals fall4Camera holds

What to write between two frames

Both ends are already pictures, so your words only cover the middle. Say what they do, what changes on the way, what drifts past, and keep the camera where both photos were taken.

“They take one step down, and their clothes change into wedding outfits, petals fall, camera stays still.”

The moveOne action that carries them from the first frame to the last.
The changeWhat is different at the end, said in a few words.
What drifts pastPetals, snow, confetti: something in the air so the middle is not just a fade.
CameraKeep it still. Both frames came from the same spot, and a camera move fights them.
in one smooth spinpiece by piecesnow fallssmooth time-lapse

Which model for which pair

Seedance 2 Mini is the default. Switch when the change between your frames is big, slow or needs sound.

Compare all 12 models
Your framesPickWhyCredits
Furniture appears piece by piece until the room is styled, smooth time-lapse, camera stays still.Any pair, first trySeedance 2 MiniCheap and lands cleanly on the last frame30 · 480P 5s
She spins once and her outfit and hair change mid-turn, smooth transition, camera stays still.A person changes lookSeedance 2Keeps the face while clothes and hair change95 · 480P 5s
The front of a small brick house with a green door in the daytime, a bare tree and a short path, no people.Slow change, 8 secondsVeo 3.1 FastMore time between frames, real light40 · 720P 8s
An empty rental living room with white walls, oak floor and a large window, afternoon light, no people.Sound for the leastMiniMax H3The lowest-cost frames run with audio15 · 480P 5s
A couple in their thirties, a Vietnamese man with short black hair and a white woman with a blonde bob, in jeans and T-shirts standing on church steps.Big moves between framesKling 4.0Handles a lot of motion without warping180 · 720P 5s

Before you add the frames: a 4-point check

Most clips that warp start from a pair that does not match. Fix the pair and the motion follows.

The same woman in the same pose in front of the same mirror, now in a black satin evening dress with her hair down in loose curls and gold earrings; same mirror, same light.A mirror selfie of a different woman, a white woman in her twenties with a red pixie cut, in a different bathroom, used as a mismatched end frame.
Same person at both ends

A different face in the last frame makes the model melt one person into another.

The same living room fully furnished: a linen sofa, a rug, plants, a coffee table and framed prints; same window, same light and angle.The same living room fully furnished: a linen sofa, a rug, plants, a coffee table and framed prints; same window, same light and angle.
Same spot, same lens

Shoot both from where you stand. A zoom between the frames reads as a jump.

A couple in their thirties, a Vietnamese man with short black hair and a white woman with a blonde bob, in jeans and T-shirts standing on church steps.A couple in their thirties, a Vietnamese man with short black hair and a white woman with a blonde bob, in jeans and T-shirts standing on church steps.
Both frames sharp

A soft last frame makes the end of the clip soft too.

A Latina woman in her twenties with a messy bun in an oversized grey hoodie and leggings posing in front of a tall mirror in her sunlit bedroom, full body, vertical.A Latina woman in her twenties with a messy bun in an oversized grey hoodie and leggings posing in front of a tall mirror in her sunlit bedroom, full body, vertical.
Matching light

Unless light is the change, like day to night, keep it the same in both.

Before and after the frames

Make the missing end frame, or keep going from the one you landed on.

Furniture appears piece by piece until the room is styled, smooth time-lapse, camera stays still.
Done

Your clip

A clip that ends on exactly the frame you chose.

No end frame yet?

Make it from the first

Edit the first frame into the last: new outfit, new furniture, night instead of day.

Image to Image
Longer piece

Chain the next shot

Your last frame becomes the first frame of the next clip, with no visible cut.

Sharper

Upscale

Take the finished clip up to 1080p, 2K or 4K.

Video Upscaler

First and last frame FAQ

What people ask before their first pair.

What is first and last frame to video?

You give the AI two pictures: the one the clip starts on and the one it ends on. It makes the 5 to 8 seconds of motion between them, so the clip lands exactly on the shot you chose.

How do I make the end frame?

Take a second photo from the same spot after the change, or edit the first frame on Image to Image: change the outfit, add the furniture, switch day to night. Keep the camera angle and the person the same.

Which models take a start and an end frame?

12 of them: Seedance 2 Mini, Seedance 2 Fast, Seedance 2, MiniMax H3, MiniMax H3 Max, Grok Imagine Video 1.5, Wan 3.0, Kling 4.0, Veo 3.1 Fast, Veo 3.1, Veo 3.1 Lite and Gemini Omni 1.2 Flash. Seedance 2 Mini is the default at 30 credits for a 480P 5s clip.

What should the prompt say?

Only what happens between the frames and what the camera does. Both ends are already fixed, so a short line such as “she spins and her outfit changes, camera stays still” works better than describing the pictures.

Do both frames need the same size?

Use the same shape for both, JPG, PNG or WEBP, and crop the second to match the first. The clip takes its shape from the first frame unless you pick another aspect ratio.

How much does a frames to video clip cost?

The credits show on the button before you start; the default run is 30. A new account gets 10 free credits for images, packs start at $9.90 and plans at $9.9 a month. A run that fails gives its credits back.

Other ways to make a video

Same models and credits, a different starting point.

All AI video models

Choose the last shot first.

Add your start frame here; it lands in the generator, ready for the end frame.

These two frames and one line“Day turns to night, snow falls and the lights switch on one string at a time, camera stays still.”