Grok Imagine-labelled image and video models appear in FastSubS. Decide whether you need a new image, an edit to an existing photo or a short video from a still image. These are distinct operations; a Grok label does not mean that every variant accepts image or video input.
Generate or edit an image
For a new image, describe the subject, composition and use: “Podcast cover: a microphone on a table, soft side lighting and empty space on the left for a title.” Photo editing requires an appropriate route and supported attachment. If attachments are unavailable, choose a compatible variant instead of assuming that a text link will be read as an image.
Prepare a photo-animation task
Tell the advisor: “Animate a photo, short calm camera movement, small budget.” Supply a suitable image and define one motion: a subtle head turn, background movement or a slow camera push. Avoid many complex simultaneous changes. Make sure you have rights to use people's images, and do not present generated footage as a real recording.
Duration, sound and price
Check the selected video model's controls. Longer duration, sound or quality can affect cost, but formulas and supported settings differ by route. A per-second rate is not the complete video's price. Review the final quote before confirming. Extending an existing video requires a source video and is a separate workflow from image-to-video.
Keep a completed result
Wait for status and inspect saved history. If the result exists, download or recover it instead of requesting a new generation. Use the website and available apps; API access is limited to documented methods. See available Grok Imagine variants and alternatives in the AI catalogue. This is not a subscription to the original Grok application.