Grok Video 1.5 Video Generation

Updated: 2026-06-26

This document explains how to use aijisu call grok-video-1.5. Model single-image-to-video: userprovidesoneInput imageandaprompt, the platformcreates an async video generation task.

1. Model

Model	Capability	Input image	Output resolution
grok-video-1.5	single-image-to-video	supports 1	480p, 720p, 1080p

Note:

• grok-video-1.5 Yes Model, do notand legacy grok-imagine.
• Currently only supportssingle-image-to-video, does not supporttext-to-video, multiple reference images, video input, video editingorvideo extension.
• the input image must be a URL accessible by the platform and upstream service.
• There is no separate landscape/portrait parameter currently; requires landscape passLandscape, requires portrait passPortrait.

2. Endpoint

2.1 Create Video Task

POST /v1/videos

request headers:

Authorization: Bearer sk-***
Content-Type: application/json

2.2 Query Video Task

GET /v1/videos/{task_id}

compatibilityqueryEndpoint:

GET /v1/videos/generations/{task_id}

2.3 Get Video File

GET /v1/videos/{task_id}/content

task completed, readqueryresponses video_url, /content endpoint to get Video File.

3. Parameter Reference

Parameters	Type	Required	Note
model	string	Yes	always pass grok-video-1.5
prompt	string	Yes	video generationprompt, recommended: describe the subject, action, camera, styleand
image	string	one of two required	Input image URL; and image_url one of two required
image_url	string	one of two required	Input image URL; and image one of two required
duration	integer	No	Output video duration, supports 1-15, recommended pass in
seconds	string	No	Output video duration compatibility field; prefer using duration
size	string	No	Output resolution, supports 480p, 720p, 1080p
resolution	string	No	Output resolution compatibility field; prefer using size

Field Constraints:

• image and image_url only pass.
• do not pass both images, image_urls, reference_image_urls Field.
• do not pass video, video_url, videos, video_urls video inputField.
• do not pass aspect_ratio.
• do not pass extra_body.operation editing, reference video or text-to-video capability.
• size/resolution Resolution tier, landscape or portrait; control orientation through the input image aspect ratio.
• If you pass both size and resolution, recommended value.

4. Request Examples

In the examples below, {BASE_URL} can be replaced with https://api.aijisu.cn.

4.1 use image_url

curl -X POST "{BASE_URL}/v1/videos" \
 -H "Authorization: Bearer sk-***" \
 -H "Content-Type: application/json" \
 -d '{"model": "grok-video-1.5",
 "prompt": "A cinematic close-up of a glass perfume bottle on a marble table, soft morning light, slow camera push-in, elegant product video, no text, no logo.",
 "image_url": "https://example.com/source.jpg",
 "duration": 6,
 "size": "720p"}'

4.2 use image

curl -X POST "{BASE_URL}/v1/videos" \
 -H "Authorization: Bearer sk-***" \
 -H "Content-Type: application/json" \
 -d '{"model": "grok-video-1.5",
 "prompt": "A quiet landscape photo becomes a gentle animated scene, leaves moving in the wind, soft natural lighting, stable camera, no text.",
 "image": "https://example.com/landscape.png",
 "duration": 8,
 "size": "1080p"}'

4.3 480p fast

curl -X POST "{BASE_URL}/v1/videos" \
 -H "Authorization: Bearer sk-***" \
 -H "Content-Type: application/json" \
 -d '{"model": "grok-video-1.5",
 "prompt": "A simple product shot, the object slowly rotates on a clean white background, smooth motion, no text.",
 "image_url": "https://example.com/product.jpg",
 "duration": 4,
 "size": "480p"}'

5. Create Response Example

A successful create request returns a task object. At this point, the video is usually queued or generating.

{
  "id": "task_xxxxxxxxxxxxx",
  "task_id": "task_xxxxxxxxxxxxx",
  "object": "video.generation.job",
  "model": "grok-video-1.5",
  "status": "queued",
  "progress": 0,
  "created_at": 1782400000
}

Save id or task_id, for later task polling.

6. Query Task

curl "{BASE_URL}/v1/videos/task_xxxxxxxxxxxxx" \
 -H "Authorization: Bearer sk-***"

In-progress Response Example:

{
  "id": "task_xxxxxxxxxxxxx",
  "object": "video.generation.job",
  "model": "grok-video-1.5",
  "status": "in_progress",
  "progress": 35,
  "created_at": 1782400000
}

Completed Response Example:

{
  "id": "task_xxxxxxxxxxxxx",
  "object": "video.generation.job",
  "model": "grok-video-1.5",
  "status": "completed",
  "progress": 100,
  "video_url": "https://example.com/generated-video.mp4",
  "result": {
    "video_url": "https://example.com/generated-video.mp4",
    "outputs": [
      "https://example.com/generated-video.mp4"
    ]
  }
}

7. Download Video

Method 1: use the value returned by the query response directly video_url.

Method 2: use the task content endpoint to get the file stream.

curl -L "{BASE_URL}/v1/videos/task_xxxxxxxxxxxxx/content" \
 -H "Authorization: Bearer sk-***" \
 -o grok-video-1.5-output.mp4

8. Status Values

Status	Note
queued	Task created and waiting for processing
in_progress	Task is processing
completed	Task is complete and the video URL can be read
failed	taskFailed, responses error.message

Clients should poll every 3 to 10 seconds, until the task enters completed or failed.

9. Common Errors

sending the model to an image endpoint

grok-video-1.5 Yesvideo model, call /v1/videos, do not /v1/images/generations.

missing input image

Model pass image or image_url. only pass prompt.

multiple images were passed

Model supportsoneInput image. do not pass images, image_urls or reference_image_urls.

video input was passed

Modeldoes not supportvideo input. do not pass video, video_url, videos or video_urls.

invalid resolution format

size or resolution only pass:

• 480p
• 720p
• 1080p

do not pass 1920x1080, 1280x720, 4k or sizestring.

aspect_ratio controlLandscapeportrait

Model does not support aspect_ratio. submit change the input image Orientation, landscapeuseLandscapeimage, portraitusePortraitimage.

requesting edit or extension capabilities

grok-video-1.5 does not supporteditingand. do not pass:

{
  "extra_body": {
    "operation": "edit_video"
  }
}

or:

{
  "extra_body": {
    "operation": "video_extend"
  }
}

10. Prompt Guidance

Prompt Guidance:

• subject: object.
• action: subject.
• camera:.
• style:.
• constraints: no text, no logo, no people.

Example:

A cinematic product video of the object slowly rotating on a clean studio background, soft natural light, smooth camera push-in, realistic texture, no text, no logo.

11. Minimum Valid Request

{
  "model": "grok-video-1.5",
  "prompt": "A calm product video, slow camera movement, soft light, no text.",
  "image_url": "https://example.com/source.jpg",
  "duration": 6,
  "size": "720p"
}