Hand-drawn anime style video from a photo

What it is

This n8n workflow turns a photo and a line about the action into a short video in a hand-drawn anime look. Fourteen nodes run in one chain with a split in the middle. A form takes the photo, shrinks it if it is larger than 1920 by 1080, and OpenAI’s gpt-image-1 redraws it in the style. The new frame goes two ways: to Google Drive, and to a vision model that describes it. A prompt writer joins that description to your idea as one Veo 3 prompt, and Veo 3 Fast on Kie.ai renders the video from the Drive copy of the frame. The last node downloads the clip.

1. A form takes a photo and a line about the action
2. OpenAI redraws the photo in the anime style
3. Drive keeps the frame while OpenAI describes it
4. A prompt writer turns both into a Veo 3 prompt
5. Veo 3 Fast on Kie.ai renders and the clip downloads

Who it suits

Creators and illustrators who want a photo to move in a painted, storybook way, and small teams that make quiet, atmospheric clips for social posts. The look sits in two prompts, so a different style means editing them. Skip it when the clip must show a real product exactly as photographed. The first step redraws everything.

Where to use it

Any n8n that can reach OpenAI, Google Drive and Kie.ai will run it. The form is the front door, so you share its address with whoever supplies the photos.

• n8n Cloud: import the file into a new workflow there
• Your own n8n server: the same file, the same steps
• Google Drive: every redrawn frame lands in one folder, so you also collect a set of styled images
• A drive or a chat: add a node after Get Veo3 Video to save the clip or send it on

Setup

Get it saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. Its author puts setup at about 13 minutes.

1. In n8n, create a workflow, open its menu, choose Import from File and pick workflow.json.
2. Add an OpenAI credential and select it in Ghibli Styler, Analyze image and veo3 Prompt Enhancer.
3. Add a Google Drive OAuth2 credential and select it in Upload file. In the same node, replace YOUR_FOLDER_ID with your folder’s ID, the part of its address after /folders/.
4. Share that folder so anyone with the link can view it, because Kie.ai opens the frame from its Drive address.
5. In Generate Veo3 Video and Get Veo3 Video Details, replace <YOUR_KIE_API_KEY> after Bearer with your Kie.ai key. The brackets go too.
6. Save, send a test photo through the form’s test address, then press Publish.

How to use it

Open the form from On form submission, attach a .jpg, .jpeg or .png photo and write the action you want in Video Description. Both fields are required. The form confirms at once. The run goes on behind it: resize, redraw, upload, description, prompt, render, download. The clip lands in the run’s execution data as a file, so open the run under Executions to see it.

Example

A photo of a harbor at dusk, with this in Video Description:

The boats rock slowly while the lamps flicker on

The redraw comes back as a 1536 by 1024 picture and goes to Drive as image-<date and time>.png. The prompt writer’s instructions ask for one flowing paragraph that opens on the described frame and ends with the words No subtitles. Kie.ai then gets a body of this shape:

{
  "prompt": "<the prompt writer’s paragraph>",
  "imageUrls": ["<the Drive address of the redrawn frame>"],
  "model": "veo3_fast",
  "aspectRatio": "16:9"
}

For newcomers and experts

If you are new. n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. A credential is a key n8n keeps for you and hands to the node you pick it in. Send the form from its test address first. Every node then shows what it received and what it sent, which is where a wrong key or an unshared folder turns up.

For experts. Two OpenAI models and Kie.ai’s Veo do the work, and two nodes decide the order.

Merge combines the two branches, so the render starts only when both the Drive upload and the prompt are done. After that, checkVideoStatus sends a data.successFlag of 1 to the download and anything else back through Wait.

Good to know

Each run costs money, and the status loop never gives up.

• Billing: OpenAI and Kie.ai each bill their own part of a run
• No limit on the loop: a task that Kie.ai fails also goes back to Wait, so it is asked about every five seconds until you stop the run
• Sharing: Kie.ai reads the frame from the file’s webContentLink, so a folder that is not shared by link leaves the render with no image
• Open form: once the workflow is published, anyone with the form’s address can start a run on your keys, so turn on the form trigger’s basic authentication or keep the address to yourself
• Your photo: it goes to OpenAI for the redraw, and the redrawn frame then goes to OpenAI again, to Google Drive and to Kie.ai
• The style: the prompts name Studio Ghibli as the look. This listing has no tie to that studio, so check how you may use the results before you publish them
• The model name: chatgpt-4o-latest is a moving alias that OpenAI may retire, so pick another vision model in both nodes if one says it is gone
• Credentials: the JSON holds none, and nothing from its author’s account