Skip to main content

​
Overview

You can access all the SOTA video and image models directly inside of Yarn, including Veo, Seeddance 2.0, and Kling. We also support image models like Nanobanana, OpenAI Image 1.0, and Seeddream. [Overview video showing:
  • Making an image
  • Making a video and showing the model picker
  • Combining images]

​
How to Generate Images and Videos

Type in the agent something like “Make me an image of a sunflower” or “Make me a subway scene video with two people talking”. The agent will switch into generate media mode. You can just press enter to [show a very short video of asking the agent to generate an image of a sunflower] Media generations can take time to complete, so in all media generations in Yarn can run in parallel.

​
Model Picker

By default, Yarn completes your request across a mixture of models. Specify this model mix by opening the model picker, where you can also set a default. By default, all generations happen with “auto” orientiation (i.e. landscape vs. Portrait)

​
How to edit images

You can edit an image by selecting it and asking the agent. “Edit this to make

​
How to combine images

You can combine images by selecting them and asking the agent “Combine these into a single image”

​
How to create videos from images

Drafting with images and then generating videos from them is

​
Advanced: Converting Videos to an overlay background

When you generate videos, by default they become clips on your timeline. You sometimes might want to convert these into overlay backgrounds. For example, if you want a generative gradient example that you want part of a reusable overlay, then you would want to convert the video into an overlay clip background. You can do this by right clicking

​
Converting Videos to a talk track

If you have a dialog scene, you might want to convert it into a talk track, for example this allows you to show captions, it allows you to control the camera position, zoom in on the phase, edit dialogue via the talk-track panel, apply persona voices.

​
Tips for Media Gen

​
FAQs

How are generative media items charged Do I pay for generations that are incomplete