Turn an image into a video in Claude and ChatGPT

Yes. Connect Makemancer and Claude, ChatGPT or Codex can turn a still image into a short video clip with sound. Send the image, describe the motion, and pick Seedance, Wan, Gemini Omni, Grok Imagine or Kling. A 5-second 480p clip with Seedance 2.0 Mini costs $0.15, quoted before it runs.

Bring
One image and a sentence about the motion
Get
An MP4 clip, with sound on most models
From
15 credits · $0.15

Made with it

Our own results, first takes. Each caption says what it took.

Fig. 1The desert motel still turned into a slow push-in with ambient soundSeedance 2.0 Mini · 480p · 5 s · 16:9 · 15 credits
Fig. 2The perfume still turned into a slow vertical orbit as the light shiftsSeedance 2.0 Mini · 480p · 5 s · 9:16 · 15 credits
Fig. 3The desert motel still animated by Wan 3.0: a car's headlights approachWan 3.0 · 480p · 16:9 · 33 credits

How to do it

  1. Connect Makemancer

    Add Makemancer to Claude, ChatGPT, Claude Code or Codex. The steps for each client are on the connection page.

  2. Send the image and describe the motion

    Share a link or attach the file. Describe what moves and what the camera does, and keep the scene the same as the still. Start with a cheap 480p draft.

    Here is my image: <link or file>. Use it as the first frame and make a 5-second 16:9 video: a slow cinematic push-in, the neon sign flickers softly, a distant car passes with its headlights sweeping the road. Use Seedance 2.0 Mini at 480p and tell me the cost first.
  3. Adjust the motion, then make the final

    If the draft moves too much or too little, change one thing in the prompt. When the motion is right, run it again at a higher resolution or on a premium model.

    The camera move is good but the car is too fast. Make the car slower and keep everything else. When I approve it, make the final at 720p.

Which model to use

ModelBest forPriced forCredits
Seedance 2.0 Minivideo/seedance-2.0-miniThe lowest-cost Seedance tier, 4-15 s at 480p or 720p with sound. Good for drafts.480p · 5 s15$0.15
Seedance 2.5video/seedance-2.5Premium cinematic clips, 4-30 s, with first and last frames.Opens after your first credit purchase480p · 5 s93$0.93
Wan 3.0video/wan-3.0Steady framing, first and last frames, 2-30 s from 480p to 1080p.Opens after your first credit purchase480p · 5 s33$0.33
Gemini Omni 1.1 Flashvideo/gemini-omni-flashThe default video model: fast, with native audio and first-last-frame support.720p · 4 s40$0.40
Grok Imagine 1.5video/grok-imagine-1.5One image as the first frame or a reference, 1-15 s at 480p or 720p.Opens after your first credit purchase480p · 8 s137$1.37
Kling 4.0video/kling-4.0-omniFirst, middle and last frames, 3-30 s at 720p or 1080p.Opens after your first credit purchase720p · 5 s157$1.57

Tips for steady motion

  • Describe the camera and one or two moving things. "Slow push-in, the sign flickers, one car passes" works. A list of ten actions makes the scene drift.
  • Say what must stay. "The bottle and label stay unchanged" or "no people" keeps the model from adding things.
  • Match the shape to the image. Ask for the same aspect ratio as your still, or the model will crop or pad it.
  • Draft cheap, finish once. Try the motion at 480p first; only the take you keep needs 720p or 1080p.

Settings that matter

Duration and resolution set the price, so state both: a longer clip costs more, and 1080p costs more than 480p. If you need the clip to end on a specific picture, give a last frame too. The video model comparison shows the same motel still animated by Seedance 2.0 Mini, Wan 3.0 and Grok Imagine 1.5.

Limits

Each generation is one continuous shot. For a sequence of shots, ask your assistant to plan them and generate each clip from its own still, then join them. Faces and hands can change during fast movement, so keep the action slow when a person is in frame.

Questions

Can Claude turn an image into a video?

Not by itself. With Makemancer connected, Claude sends your image to a video model such as Seedance or Wan and returns the clip with a permanent link.

Can I animate an image in ChatGPT?

Yes, if your ChatGPT plan and workspace allow custom MCP servers. Add Makemancer, attach the image and describe the motion; the /how-to-connect page has the setup steps.

How long can the video be?

It depends on the model: Seedance 2.0 Mini makes 4-15 s, Grok Imagine 1.5 makes 1-15 s, and Seedance 2.5, Wan 3.0 and Kling 4.0 go up to 30 s. Longer clips cost more.

Does the video have sound?

Most models here generate audio with the picture, including Seedance, Wan, Gemini Omni, Grok Imagine and Kling. Describe the sound in the prompt if it matters.

Can I set both the first and the last frame?

Yes, with Gemini Omni Flash, Seedance 2.0 Mini or MiniMax H3, and after your first credit purchase also with Seedance 2.5, Wan 3.0 or Kling 4.0. Send two images and say which one opens and which one closes the clip.

How much does it cost to animate a photo?

From $0.15 depending on model, length and resolution. A 5-second 480p Seedance 2.0 Mini clip costs $0.15; the assistant quotes every request first.