PhotoLoop

AI hug video generator from your own photos

Upload a real two-person photo or two portraits, preview the prepared frame, then generate an experimental hug video from it.

Prepare a hug starting frame

Use one existing two-person photo, or choose two portraits. Two portraits are combined locally in your browser into one video reference image.

Use photos you own or are authorized to use, and make sure the people pictured consent to this use.

This two-portrait workflow is experimental. Cropping and the join are not guaranteed to be seamless; an existing two-person photo is usually more reliable, and identity is not guaranteed.

214/500
Need a starting point? Browse prompts by scene
Video

Your photo and prompt are sent to our model provider; free-tier photos and clips are deleted from our servers within 48 hours, and we never train on them.Loading balance…

Upload, preview, then generate

Use an existing photo of two consenting people when you have one. If you only have separate portraits, PhotoLoop combines them in your browser and shows the prepared frame before you decide to generate.

Keep the motion brief

The prepared reference uses a short, steady hug direction. You can edit the prompt in the generator before submitting, but a single continuous motion is easier to judge than a multi-scene request.

Review the result as an experiment

Two-portrait composition, cropping, and generated motion can drift. An existing two-person photo is usually a better starting point, and no workflow can guarantee identity, expressions, or a seamless join.

  • JPG, PNG, or WebP up to 15 MB per source image
  • Two portraits are combined locally to a bounded 16:9 JPEG
  • The prepared image is sent only when you choose Generate
  • Use only photos you own or are authorized to use, with consent from pictured people

Hug video questions

Does PhotoLoop upload both portraits while combining them?

No. The two portrait files are combined in your browser. Only the prepared image is sent when you choose to generate a video.

Will the people remain identical in the generated video?

Identity and motion are model outputs and are not guaranteed. Compare the result with your source and avoid using photos or prompts that imply consent you do not have.

Is the two-photo result seamless?

It is an experimental side-by-side preparation. Cropping and the join may be visible, so an existing two-person photo is usually the more reliable input.

How does the free use work?

The shared free video allowance is one preview per device per day. Previewing and combining photos locally is free; an account is needed to download a generated video.

Do I need an account to preview?

No. You can upload, combine, preview, and submit the starting frame first. Sign in when you want to download the generated video.

Keep exploring

PHOTOLOOP