grok-imagine-video:reverse is the video generation entry point for the xAI Grok Imagine series, supporting both video creation from text descriptions and dynamic shot design starting from static images. It is suitable for producing 1–15 second creative clips, product showcases, and storyboard previews. When creating, you can organize inputs around the subject, action, scene, and camera movement; after generation, obtain a video link and integrate it into editing or content publishing workflows.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
Input parameters and result formats vary by service. Use the public API for this model and follow its guide for generation, task retrieval and editing operations.
The durations and operating methods above apply to this API endpoint and do not treat the native specifications of other versions in the series as guarantees for this model.
When no ready-made image is available, you can directly describe the subject, environment, action, and camera changes to generate a video. Prompts work well as a clear single shot—for example, how the subject moves through a space, how the camera approaches, and what atmosphere the lighting creates. Keeping the focus on one action makes it easier to compare different creative options.
When you already have a product image, character image, or scene image, you can enter the image-to-video workflow through image_url and then describe the action you want to occur in text. The image serves as the visual starting point, while the prompt adds motion intent. This is suitable for exploring short segments such as a character turning, environmental changes, or camera push-ins, rather than redefining the entire image from text.
Video generation can be submitted asynchronously: first obtain a task_id, then query the status or receive a completion callback. Results provide video_url, allowing applications to handle submission, waiting, previewing, and downloading separately. For content production systems, this task-based delivery is easier to organize than requiring users to remain on a waiting page throughout the process.
Use static product images as input, add background atmosphere, motion direction, and camera distance, and generate candidate short videos for presentation. Start by trying one clear action, then pass suitable clips to the editing stage to add selling-point text, logos, and music. Suitable for teams that need to expand dynamic content from existing visual assets.
Rewrite a shot from the script as descriptions of the subject, scene, action, and camera movement, then create a watchable draft through text-to-video. Deliverables are used to discuss pacing, composition, and action intent, without needing to prepare complete live-action footage early on; different descriptions can be generated separately to help creators select a more suitable direction of expression.
Use existing illustrations or character images as a starting point, specify subtle movements, viewpoint changes, or environmental motion, and create dynamic proposals. Suitable for exploring presentation formats for event visuals, character showcases, and social content. The details that need to be retained in the final video can be specified in the prompt and checked item by item when selecting clips for subject form and motion results.
If you need to retain both text-to-video and image-to-video modes, and each clip is within the 1–15 second range, choose this model. grok-imagine-video-1.5-fast:reverse has a duration range of 6–30 seconds, making it more suitable for tasks requiring longer clips. They are not the same invocation ID; when switching, adjust the duration settings accordingly rather than merely replacing the name while keeping all parameters.
grok-imagine-video:official also provides text-to-video and image-to-video modes and is another public invocation endpoint; grok-imagine-video-1.5:official supports image-to-video only and requires an image. When you only have a text concept, this model can be used directly; when you already have images and are ready to compare other versions, you can use the same assets and action descriptions to test each one separately.
For text-to-video, provide a prompt; for image-to-video, use image_url to define the visual starting point and add motion instructions. Configure reference image capabilities according to this public variant guide.
Specify model=grok-imagine-video:reverse for /grok/videos, starting with a small 6-second task; use durations of 1–15 seconds. Choose aspect_ratio and resolution based on the actual footage.
Use async or callback_url to track the task, and use task_id to query /grok/tasks; wait for succeeded before reading video_url, check the subject, motion, visual continuity, and ending, then move the returned video into editing as appropriate.
Use a photo of greenery by a window as the first frame, with the leaves gently swaying and the camera slowly pushing in, while preserving the flowerpot, window frame, and composition.
Provide image_url and clearly specify subtle motion, selecting 1–15 seconds; do not apply the 6–30 second range of fast:reverse.
Yes. This model supports text-to-video; simply provide a prompt to describe the content you want to generate. It is recommended to specify the subject, scene, action, and camera movement, organizing the information around a single shot first. When calling it, explicitly specify grok-imagine-video:reverse to avoid using other default models.
When providing image_url, prompt can be left blank. However, if you want the subject to move in a specific direction, or want the camera to push in or rotate, adding an action description makes your intent easier to express. The image provides the visual starting point, while the text explains how you want the scene to change; they serve different roles.
This model supports durations of 1–15 seconds, with a default of 6 seconds, and is not suitable for directly submitting a 30-second task. For longer single-shot videos, consider grok-imagine-video-1.5-fast:reverse, which supports durations of 6–30 seconds; for multiple shots, you can also generate them separately and edit them together.
task_id is a task identifier, not a video URL. When using async, you can check progress through POST /grok/tasks; when callback_url is set, you can wait for the completion notification. After confirming that the result status is succeeded, read video_url; pending indicates that processing is still in progress.
:reverse is a suffix for the public call ID, used to distinguish endpoints, and should not be treated as an independent vendor model name. When using this endpoint, retain the complete ID. It shares the basic text-to-video and image-to-video creation methods with the :official endpoint, but this does not mean that all optional parameters are exactly the same.
Choose a development approach or application suited to this model and follow its setup guide.
↗Python SDK Integration Guide
SDK installation, authentication and examples↗TypeScript SDK Integration Guide
SDK installation, authentication and examples
↗Pricing
$0.0599· per generation
You pay$13,584for 142,800 credits · $0.0951 / credits46% OFF
Estimated price
You're billed in credits. Pick a top-up tier to see its per-credit rate — the larger the top-up, the cheaper each credit.
This is only a basic call example. See the full docs for more parameters and advanced usage.
View Docs