Skip to content

Photo to video

Upload a photo, describe the movement, and create a short video.

Generated in Ofrai · GPT Image 2.5 → Seedance 1.0 Pro Fast · 5 seconds

From an idea to a moving story

Try photo to video ↑
Space cat courier

Original fictional character. This example used 21 image credits and 41 video credits. Current prices appear before generation.

Watch the helmet, orange hover scooter, and headlight. All three stay recognizable for five seconds. Road reflections and the shifting background carry much of the motion. For your own photo, choose one thing to move and two details to check.

Creative demos

A new story from one photo

Explore an idea, then make it your own.

Step into a paper moonlit city

A garden dress in motion

Jellyfish beyond the rain window

Creative demos. Results vary with your inputs and settings.

How could your photo move?

Bring a portrait to life

Gentle hair and fabric movement for an atmospheric portrait clip.

Try this camera move

Give a product its own shot

Slow camera movement and changing light for showcases and launch teasers.

Try this camera move

Put an outfit in motion

Try a slow walk for fashion content and outfit presentations.

Try this camera move

Clips are motion references, not verified Ofrai test results. Output varies with your photo, model, and prompt.

From photo to video in three steps

  1. Upload your photo

    Start with a clear, well-lit portrait or product photo with an easy-to-read subject.

  2. Describe the motion

    Edit the prompt directly, or start with a camera preset. Describe one action at a time.

  3. Review and generate

    Check the duration and credit estimate before submitting. Preview and download the completed clip.

See the cost before you commit

With Seedance 2.5, one photo and a 5-second 16:9 clip at 480p is quoted at 528 credits. Budget 1,056 credits for two attempts at those settings.

Start with a 5-second, 16:9 clip by default. The generation panel shows the credit estimate for the current model and settings. Each regeneration is a new job.

Keep your first move simple

Try gentle movement or a slow camera move first. Large head turns, fast walking, and occlusion can change faces, product geometry, or text. Review the full clip before publishing.

Before you start

Do I need to generate an AI image first?
No. Upload an existing photo, write a motion prompt, and submit a video job directly on this page.
Can I write my own prompt?
Yes. Edit the prompt field directly. Presets are starting points: describe the subject's action, the camera movement, and what should stay unchanged.
Will faces and product details stay identical?
Exact fidelity is not guaranteed. Begin with a sharp photo and small movements. Check faces, hands, labels, and backgrounds; simplify the action if you need to try again.
Do the reference videos show the result I will get?
No. They illustrate motion directions and are not verified Ofrai outputs of the selected model. Your result depends on your photo, model, and prompt.

Let the jellyfish move while the portrait stays still

The jellyfish and window reflections carry the motion. For a first attempt with your own portrait, ask for a small change in gaze, then check whether the window frame and shoulder line drift.

See how to break down the rain-window shot

Give your first clip a stronger start

Learn to choose a photo, adapt motion prompts, and troubleshoot changing faces or sliding backgrounds.

Read the photo-to-video guide