Skip to main content
Grok Imagine Video is xAI’s video generation family, available in ComfyUI through the Grok partner nodes. It covers video generation from text or an image, video editing, seamless video extension, and reference-to-video generation guided by character or style images. For the latest model, see Grok Imagine Video 1.5.
To use the Partner Nodes, you need to ensure that you are logged in properly and using a permitted network environment. Please refer to the Partner Nodes Overview section of the documentation to understand the specific requirements for using the Partner Nodes.
Make sure your ComfyUI is updated.Workflows in this guide can be found in the Workflow Templates. If you can’t find them in the template, your ComfyUI may be outdated.If nodes are missing when loading a workflow, possible reasons:
  1. You are not using the latest ComfyUI version (Nightly version)
  2. Some nodes failed to import at startup

​
Grok: Video Generation

Generate a short video from a text prompt, with an optional starting image. This workflow uses the grok-imagine-video model to produce up to 15-second clips with synchronized audio.

Run on Comfy Cloud

Open in Comfy Cloud

Download Workflow

Download JSON or search “Grok Imagine” in Template Library

​
Workflow overview

This workflow uses three nodes:
  • LoadImage: provides the optional starting image frame
  • GrokVideoNode: the core node, configured with the grok-imagine-video model
  • SaveVideo: saves the generated video with audio

​
Steps to run

  1. Enter your prompt: describe the motion, atmosphere, and scene dynamics in the GrokVideoNode
  2. Optionally load a starting image: connect an image to the LoadImage node
  3. Set resolution: 480p or 720p
  4. Set duration: choose a length from 1 to 15 seconds
  5. Set seed: control whether the node re-runs; outputs are not reproducible regardless of seed
  6. Click Queue: press Ctrl+Enter to generate

​
Grok: Video Edit

Edit an uploaded video with natural-language instructions. Restyle the footage, change the setting, or apply a new aesthetic while keeping the original motion and content.

Run on Comfy Cloud

Open in Comfy Cloud

Download Workflow

Download JSON or search “Grok Imagine” in Template Library

​
Workflow overview

This workflow uses three nodes:
  • LoadVideo: provides the source video to edit
  • GrokVideoEditNode: the core node, configured with the grok-imagine-video model
  • SaveVideo: saves the edited video

​
Steps to run

  1. Load a video: connect your video to the LoadVideo node (1 to 8.7 seconds, up to 50MB)
  2. Describe the edit: tell the model how to transform the footage in the GrokVideoEditNode prompt
  3. Click Queue: press Ctrl+Enter to generate

​
Grok: Video Extend

Extend a short video with a seamless continuation. Describe the next scene in natural language, and the model generates a new clip that follows on from the original footage.

Run on Comfy Cloud

Open in Comfy Cloud

Download Workflow

Download JSON or search “Grok Imagine” in Template Library

​
Workflow overview

This workflow uses three nodes:
  • LoadVideo: provides the source video to extend
  • GrokVideoExtendNode: the core node, configured with the grok-imagine-video model
  • SaveVideo: saves the extended video

​
Steps to run

  1. Load a video: connect your video to the LoadVideo node (MP4, 2 to 15 seconds, up to 50MB)
  2. Describe the next scene: tell the model what should happen next in the GrokVideoExtendNode prompt
  3. Set duration: choose the extension length (2 to 10 seconds)
  4. Click Queue: press Ctrl+Enter to generate

​
Grok: Reference to Video

Generate a video guided by up to 7 reference images. Use character, product, or style references to keep the subject consistent while the prompt describes the motion and scene.

Run on Comfy Cloud

Open in Comfy Cloud

Download Workflow

Download JSON or search “Grok Imagine” in Template Library

​
Workflow overview

This workflow uses four nodes:
  • LoadImage x2: provide the character and object reference images
  • GrokVideoReferenceNode: the core node, configured with the grok-imagine-video model and reference images
  • SaveVideo: saves the generated video

​
Steps to run

  1. Load your references: connect up to 7 reference images to the LoadImage nodes
  2. Enter your prompt: describe the scene, motion, and how the references appear in the GrokVideoReferenceNode
  3. Set resolution: 480p or 720p
  4. Set aspect ratio: pick from 16:9 down to 9:16
  5. Set duration: choose a length from 2 to 10 seconds
  6. Click Queue: press Ctrl+Enter to generate