Can I turn a photo into a video with AI?

Yes. Upload the photo, describe the motion, and Pose generates a video that starts from your image. SeedDance 2.5, SeedDance 2.0, Kling 3.0 and Gemini Omni Flash are all available.

Which AI model is best for image to video?

SeedDance 2.5 is the default and handles people and motion well, with voice. Kling 3.0 is strongest for cinematic camera moves at 1080p, and SeedDance 2.0 on its Standard tier is the route to 4K.

How long can the video be?

From 3 to 15 seconds depending on the model: up to 15 seconds on Kling 3.0 and 10 seconds on SeedDance 2.5, SeedDance 2.0 and Gemini Omni Flash when animating a photo. For longer clips from a text prompt, see text to video, where SeedDance 2.5 goes up to 30 seconds.

Will the person still look like me?

Yes. The video starts from your photo and keeps the identity in it, so the face stays recognisable through the clip.

SeedDance 2.5 and 2.0 can generate speech, including a cloned voice on SeedDance 2.0; Gemini Omni Flash generates a voice automatically from the script.

Is there a watermark?

No. Videos download without a watermark on the paid plan — $4.99 for the first week, then $14.99 a week with 400 credits.

Image to Video AI: Turn Any Photo into a Video

Upload a photo, describe what should happen, and Pose turns it into a video. A portrait becomes a short clip of you moving or speaking; a product shot becomes a slow reveal; a landscape gets a camera move. The first frame is your image, so the result starts from exactly what you chose.

Four video models are built in, each picked for a different job, and your face stays recognisable because identity is locked from the photo itself.

$4.99 first week, then $14.99/week · 400 credits every week · Cancel anytime

TL;DR

Upload one image and describe the motion; Pose generates the clip from that first frame.
Models: SeedDance 2.5 (the default, with voice), SeedDance 2.0 (up to 4K on its Standard tier), Kling 3.0 (1080p, strong camera moves) and Gemini Omni Flash (the fastest and cheapest, with an auto-generated voice).
Clips of 3 to 15 seconds depending on the model; the price updates as you change length and resolution.
Export in 16:9, 9:16 or 1:1 for YouTube, TikTok, Reels and ads.

How to turn an image into a video

1. Upload your image

A portrait, a product photo, a scene — or pick one from My photos or your Gallery. A sharp, well-lit image gives the model more to work with.

2. Describe the motion

Say what moves and how: "she turns to camera and smiles", "slow push-in on the bottle", "wind through the trees". The Enhance Prompt toggle can expand a short description for you.

3. Choose a model and length

Pose starts on SeedDance 2.5. Switch to Kling 3.0 for cinematic camera work, SeedDance 2.0 Standard for 1080p or 4K, or Gemini Omni Flash for a quick draft.

4. Generate and download

The clip lands in your Gallery, ready to download without a watermark.

Which model to pick

Model	Best for	Length	Resolution	Voice
SeedDance 2.5	Default — people, motion, talking clips	5–10 s	480p, 720p	Yes
SeedDance 2.0	The highest resolution	4–10 s	Up to 4K (Standard tier)	Yes, incl. cloned voice
Kling 3.0	Cinematic camera moves, scenes	5, 10, 15 s	1080p	—
Gemini Omni Flash	Fast, cheap drafts	3–10 s	360p to 4K	Auto-generated

Prices are shown in the app for each length and resolution before you generate.

Image to video vs Clone a Video

Image to video invents the motion from your description. If you want to copy the motion of a clip that already exists — a trending dance, a transition — use Clone a Video instead: it keeps the clip's movement and puts your face or product in it.

For a talking video from a script, see UGC talking videos; to start from a text description with no image, see text to video.

FAQ

Frequently asked questions

Animate your first photo

One image, one sentence of motion, and a clip in your Gallery.