Bring short-video generation into a production workflow
Create a dedicated Grok Video API key to submit text-to-video, image-to-video, editing, or extension jobs. Track every asynchronous task with clear specifications and estimates.
Keep generation and follow-up creation in one task flow
Video requests do not hold the browser connection open. Save the request_id, poll status, and preview or download the result when it is ready.
Text to video
Describe the scene, subject, camera motion, and pacing to create a short video.
Image to video
Upload one primary reference image and add motion, camera movement, and time.
Edit video
Create a new task from an existing video and an editing instruction.
Extend video
Continue from the end of a source video and receive a new request_id.
Use the model that matches the input
Text generation, editing, and extension use the base model. Image-to-video uses the 1.5 model, and the console prevents unsupported combinations.
| Model | Best for | Resolution | Duration |
|---|---|---|---|
| grok-imagine-video | Text, edit, extend | 480p / 720p | 1–15s |
| grok-imagine-video-1.5 | Single-image video | 480p / 720p / 1080p | 1–15s |
Submit once and track the request_id to completion
Creation endpoints accept the task. Save the returned identifier and query status because there is no upstream task-list API.
Create and use a dedicated Grok Video API key for generation, editing, or extension.
Save the returned request_id in your own task record.
Poll status and preserve the identifier when an unknown status appears.
Stream or download the completed result before it expires.
Use the /v1 paths. Status and content requests must use the same API key that created the task.
curl https://api.gettoken.dev/v1/videos/generations \
-H "Authorization: Bearer $GETTOKEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "grok-imagine-video",
"prompt": "A quiet city street after rain",
"resolution": "720p",
"duration": 8
}'Estimate before submission and trust actual usage afterward
The server quotes model, resolution, and duration before submission. Actual cost appears only when the status result explicitly includes it; otherwise the value remains an estimate.
Users do not select or switch the video pricing group.
Completed results remain available for 24 hours.
HTTP Range support enables streaming and timeline seeking.
Cancellation, deletion, batch generation, and multiple reference images are not available yet.
FAQ
Start with one controlled video task
Choose an existing API key, review the model, resolution, duration, and quote, then complete the first generation in the workspace.
Continue with GetToken
Compare plans, follow the setup guide, or check product and support details.