This visual effect I created was a hit on Instagram. Using the agentic chat on Venice, we can do everything on the platform without the need for any external LLM or agent subscriptions. Full process and prompt template below 👇

Grab the full quality prompt template here:

1. Go to @AskVenice to use the Agentic Chat function.

2. Drop in this PDF containing the prompt template and the image you want to animate. This is the image I used for this particular video & effect.

3. Tell the agent exactly what you want to see: "Following the instructions on this document, write me a 15-second prompt that describes this model smoking her cigarette, she exhales the smoke and lies back on the hood of the car. Her body melts into the paint of the car, as she transforms into a beautiful mural."

4. Then we copy our prompt. We take our image into Venice Studio, where we select the Seedance 2.0 R2V model )Reference 2 Video). Select your time, aspect ratio, and resolution.

venice. I have learnt a lot watching your youtube long videos. thanks

So dope! Curious why you used Agentic Chat instead of normal Chat where you can select the text model to create the prompt?

I was honestly just testing the chops of the Agentic Chat since it’s a newer feature on the site. Definitely was not disappointed and reminded me alot of working with Claude locally (which is a good thing!)

Rad seeing it in a workflow that produced such a great product. Really curious what the Agentic Chat architecture is (model, harness, etc).

VENICE

Nice Venice workflow, keeps everything in one place