How do I generate videos with Grok Imagine?
Updated August 2026
Ask an AI about this guide
Open this article in your assistant with a ready-made question about doing the same job on your own API keys.
SeamUI supports Grok Imagine Video through direct xAI and Kie.ai routes. Start from a text prompt or an approved image, choose only the inputs exposed for that route, and keep the finished clip beside its source assets.
Generate a Grok video in SeamUI
Open a project, switch the generation sidebar to Video, then choose Direct xAI or Kie.ai and a Grok Imagine Video model. Write the scene and motion prompt, or attach an image when the selected route exposes image-to-video inputs.
Some Grok routes also accept reference images or a supported source-video task for editing or extension. SeamUI reads those inputs from the same capability registry as the generator, so unavailable combinations are not presented as universal features.
- Describe camera movement, subject action and pacing explicitly.
- Use an approved still when visual consistency matters more than open-ended exploration.
- Review the selected route's input and setting labels before submitting.
Honest workflow limits
Duration, resolution, editing and input support vary by model and route. SeamUI shows the controls that the chosen route supports instead of applying one capability claim to every Grok video model.
Video is not batch generated. Each request creates one video job. Image CSV workflows and batches of up to 10 remain image-only.
Direct xAI or Kie.ai
Use a direct xAI key when you want the xAI route, or a Kie.ai key when you prefer one aggregator connection. In both cases the provider bills generation usage directly, with no SeamUI markup.