Instagram profile animation from a photo and a screenshot

What it is

This n8n workflow makes a short vertical video from two images: a photo of you and a screenshot of an Instagram profile. A form takes both, and Cloudinary stores them at public addresses. Kie.ai’s Nano Banana Edit then composes a 9:16 picture of the person stepping out of a glowing glass Instagram screen that shows the profile’s details. Veo 3 Fast on Kie.ai animates that picture with a slow push-in and floating hearts and comments. Last, WaveSpeedAI’s video face swap puts the face from your original photo onto the person in the clip, and the final node downloads the video. Each service is polled until it reports a result.

Who it suits

Creators and small brands that want one striking vertical clip of themselves for Reels, Shorts or TikTok, and the social media managers who make it for them. It works with your own face and your own profile, or with someone who has agreed to both.

You need some n8n: importing a file, pasting keys into six nodes and swapping a few placeholders. If that is new and you would rather not learn it, skip this one. The page it came from estimates 11 minutes, which assumes the three accounts already exist.

Where to use it

Any n8n you run will do, as long as it can reach Cloudinary, Kie.ai and WaveSpeedAI.

• n8n Cloud: import the file into a new workflow there
• Your own n8n server: the same file, the same steps
• Instagram Reels, TikTok and YouTube Shorts: the clip is 9:16, so it fits all three once you have saved the file
• A video editor: trim it or add a caption before you post

Setup

Get it saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. Have these ready before you import it.

1. In n8n, create a workflow, open its menu, choose Import from File and pick workflow.json.
2. In Upload Photo and Upload Screenshot, put your cloud name in the address where it says <YOUR_CLOUD_NAME>. Swap YOUR_PRESET_NAME for your preset and YOUR_FOLDER_NAME for a folder.
3. In the Authorization header of nano-banana-edit, nano-banana-result, veo3-video and veo3-video-status, swap <YOUR_KIE_KEY> for your Kie.ai key and keep the word Bearer and its space.
4. Do the same in video-faceswap-submit and video-faceswap-result with <YOUR_WAVESPEED_KEY>.
5. Save, send a photo and a profile screenshot through the form’s test address, then press Publish.

The page it came from tells you to create credentials and pick them in an Authentication field. The file holds no such setting, so the keys go in the header fields. If you would rather keep them in credentials, create an HTTP Header Auth credential for each service, select it in those nodes and delete their Authorization header.

How to use it

Open the form from On form submission and attach two images, each a jpg, jpeg or png. Both are required: a clear photo of the person, and a screenshot of the profile to show. The form confirms at once and the run goes on behind it: image, then video, then face swap, each checked every few seconds until it is done. The finished video is downloaded by the last node, final result, and sits in the run’s data. Open the run under Executions to see it.

Example

The two images sent through the form:

Upload your photo: portrait.jpg
Upload your screenshot: profile-screenshot.png

Nano Banana Edit gets both pictures and its prompt, which asks to keep the face and to copy the screenshot’s username, bio, posts and follower counts into the glass screen. Veo 3 Fast animates the result. WaveSpeedAI then receives this request, which puts your original face on the person in the clip:

{
  "face_image": "<your photo’s Cloudinary address>",
  "max_duration": 0,
  "target_gender": "all",
  "target_index": 0,
  "video": "<the Veo clip’s address>"
}

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. Here three services take a while to finish, so the workflow waits, asks how the job is going and loops until the answer is done. Run the form from its test address first. Open that run under Executions and each node shows what it received and what it sent, which is where a wrong key or a failed task turns up.

For experts. Each of the three services reports its progress in its own words, and a Switch node reads it.

A failed job ends in a No Operation node, so the run finishes without an error and without a clip. None of the three loops has a retry cap. Set a workflow timeout in n8n’s settings if you want an upper bound.

Good to know

Each run costs money, and the clip is a face swap, so mind whose face it is.

• Billing: Kie.ai bills the picture and the video, and WaveSpeedAI the face swap
• Whose face: use your own face and a profile you run, or ones whose owner agreed. A swapped face in a public clip is not something to do to someone unasked
• Fixed look: the prompts describe a young man in a white hoodie, black shorts and sneakers, so edit both prompts for anyone else
• Quiet failures: a failed job ends the run without an error, so check the last node that ran
• Open form: once it 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 private
• The images: both uploads become public Cloudinary addresses and go to Kie.ai and WaveSpeedAI, and a profile screenshot shows a name, a bio and numbers
• Keys: they sit in header fields of six nodes, so they travel with any export of the workflow
• A repair: the page says the three waits are 30, 30 and 10 seconds, but the file gave them no unit and n8n would fill in its own default. Mindket wrote seconds into each
• Credentials: the JSON holds none, and nothing from its author’s account