YouTube 'Make me move' and Photo to video: limits and a Sume route

YouTube's Photo to video turns one of your photos into an animated Short, and Make me move animates a photo of yourself with a preset movement style. YouTube says standalone AI clips in Shorts are up to 10 seconds long. If you cannot see either tool, or you want a clip you control, Sume's video API can animate a still from a first-frame image, with no YouTube account involved.

The YouTube facts below come from Create content for Shorts using AI-generated features and Create YouTube Shorts, both read on 2026-10-02. Sume's side comes from Video Generation.

What do Photo to video and Make me move do?

YouTube lists five AI-generated features for Shorts. Two of them start from a photo: Photo to video transforms personal photos into animated Shorts with suggested prompts, and Make me move animates a photo of yourself using preset movement styles. The others are AI green screen backgrounds, standalone video clips of up to 10 seconds from text prompts and reference images, and audio-first conversion of podcasts.

YouTube also says it incorporates safeguards to prevent generating photorealistic images of identifiable people that could be used for deepfakes, and that content must follow Community Guidelines.

Where are these tools missing?

The availability list on YouTube's page is the first thing to check before you plan a workflow around them.

• Audio-first conversion is a separate tool, limited to Australia, Canada, the UK and the US, on desktop only.

Do you have to label a Short made with these tools?

YouTube says it discloses AI tool usage automatically, so creators do not need to take extra steps for content made with these tools. That only covers clips made inside YouTube. A clip you generate elsewhere and upload is your responsibility to disclose, and our post on what counts as altered or synthetic content covers when a label applies.

How do you animate a photo with Sume instead?

Sume's POST /v1/videos accepts frame_images for image-to-video, where a first_frame image is the opening shot. Set aspect_ratio to 9:16 and a duration in whole seconds, and use model: "sume/auto" to let Sume pick the family. The image URL must be public HTTPS.

Send an Idempotency-Key so a retry returns the original job instead of a second charge, then poll the polling_url and download unsigned_urls[0] when the status is completed.

curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: photo-to-short-001" \
  -d '{
    "model": "sume/auto",
    "prompt": "Slow push-in, hair moves in a light breeze, subject smiles",
    "aspect_ratio": "9:16",
    "duration": 5,
    "frame_images": [{
      "type": "image_url",
      "image_url": {"url": "https://example.com/photo.png"},
      "frame_type": "first_frame"
    }]
  }'

What does Sume not do here?

Sume has no preset movement styles and no Make me move button; you describe the motion in the prompt and the result is not guaranteed to match. Sume does not publish to YouTube, so the clip is a file on media.sume.com that you upload yourself in Studio. Durations are per model: check GET /v1/videos/models for each model's supported_durations and supported_aspect_ratios before you submit.

Billing is a workspace USD balance reserved on submit at provider list price times 1.25, per the docs. If the clip runs longer than you want, Video trim cuts a range into a new MP4 for $0.02 per job.