Grok Imagine Video 1.5 is an xAI model for generating videos from text descriptions and visual material. Ailoot's current method publishes generation from text; additional modes of the model itself allow image animation, use of references and specification of anchor frames in a scene.
The result is created asynchronously: after a task is submitted, readiness is checked through its identifier. According to xAI, the process may take several minutes depending on scene complexity, duration and resolution. Receiving a task identifier does not yet mean the video is ready.
Generated videos last 1 to 15 seconds. 1:1, 16:9, 9:16, 4:3, 3:4, 3:2 or 2:3 can be selected; the standard format for text is 16:9. Image animation retains the image's proportions by default.
480p, 720p and 1080p are available for generation from text and a starting image. 480p is the default; reference mode is limited to 720p. The existence of 1080p in the model does not mean it applies to every mode.
In reference mode, xAI accepts up to 7 images. They define people, objects, clothing and other visual characteristics without fixing the initial frame. This method is intended for scenes with a recurring character, product demonstrations and retention of the desired appearance across different stagings.
Version 1.5 allows the exact first frame, exact last frame or both to be fixed. If only the last is specified, the model constructs the beginning itself and arrives at the supplied image. With both boundaries specified, it creates movement between them; a text instruction can refine motion and camera work.
For storyboarding within a video, up to 4 keyframes can be fixed. Each time must be strictly after the start and before the end; placement follows a 1/3-second grid. Two frames that round to the same grid point are rejected. To avoid this, leave at least 1/3 second between them. First and last frames are specified separately from these four intermediate frames.
Anchor frames can be combined with visual references and preset voices. If at least one frame is fixed, a text description is optional: the image sequence itself defines the scene. Keyframes cannot be used with video-editing mode. These xAI capabilities do not mean they are available in Ailoot's published text-based generation.
A generated video contains an audio track by default; a silent result is also provided. In xAI's voice-reference mode, up to 3 preset voices can be selected and associated with speaking participants in the scene.
Supplying a custom audio file as a voice sample is not a general capability of this model: xAI provides it separately on request. A preset voice and uploaded audio are different types of input.
Text-to-video generation creates a new scene; a starting image defines its beginning; references guide appearance; fixed frames specify visual points in time. Each mode has its own limitations, so settings from one method cannot be unconditionally applied to another.
", "aboutText": "### Creating a video scene\n\nGrok Imagine Video 1.5 is an xAI model for generating videos from text descriptions and visual material. Ailoot's current method publishes generation from text; additional modes of the model itself allow image animation, use of references and specification of anchor frames in a scene.\n\nThe result is created asynchronously: after a task is submitted, readiness is checked through its identifier. According to xAI, the process may take several minutes depending on scene complexity, duration and resolution. Receiving a task identifier does not yet mean the video is ready.\n\n### Duration, proportions and resolution\n\nGenerated videos last **1 to 15 seconds**. `1:1`, `16:9`, `9:16`, `4:3`, `3:4`, `3:2` or `2:3` can be selected; the standard format for text is `16:9`. Image animation retains the image's proportions by default.\n\n**480p, 720p and 1080p** are available for generation from text and a starting image. 480p is the default; reference mode is limited to 720p. The existence of 1080p in the model does not mean it applies to every mode.\n\n### Controlling appearance\n\nIn reference mode, xAI accepts up to **7 images**. They define people, objects, clothing and other visual characteristics without fixing the initial frame. This method is intended for scenes with a recurring character, product demonstrations and retention of the desired appearance across different stagings.\n\n### First, last and intermediate frames\n\nVersion 1.5 allows the exact first frame, exact last frame or both to be fixed. If only the last is specified, the model constructs the beginning itself and arrives at the supplied image. With both boundaries specified, it creates movement between them; a text instruction can refine motion and camera work.\n\nFor storyboarding within a video, up to **4 keyframes** can be fixed. Each time must be strictly after the start and before the end; placement follows a **1/3-second** grid. Two frames that round to the same grid point are rejected. To avoid this, leave at least 1/3 second between them. First and last frames are specified separately from these four intermediate frames.\n\nAnchor frames can be combined with visual references and preset voices. If at least one frame is fixed, a text description is optional: the image sequence itself defines the scene. Keyframes cannot be used with video-editing mode. These xAI capabilities do not mean they are available in Ailoot's published text-based generation.\n\n### Audio and voices\n\nA generated video contains an audio track by default; a silent result is also provided. In xAI's voice-reference mode, up to **3 preset voices** can be selected and associated with speaking participants in the scene.\n\nSupplying a custom audio file as a voice sample is not a general capability of this model: xAI provides it separately on request. A preset voice and uploaded audio are different types of input.\n\n### Operating modes\n\nText-to-video generation creates a new scene; a starting image defines its beginning; references guide appearance; fixed frames specify visual points in time. Each mode has its own limitations, so settings from one method cannot be unconditionally applied to another.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "318189dd18033cbc3efff06e2b986903e591b8b11c9e5fe3bc31e2219063e1a1", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "grok-imagine-video-1.5-previewGrok Imagine Video 1.5 creates 1–15-second videos from text, including 1080p. The model itself supports image animation, visual references, keyframe staging and an audio track.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Grok Imagine Video 1.5 is an xAI model for generating videos from text descriptions and visual material. Ailoot's current method publishes generation from text; additional modes of the model itself allow image animation, use of references and specification of anchor frames in a scene.
The result is created asynchronously: after a task is submitted, readiness is checked through its identifier. According to xAI, the process may take several minutes depending on scene complexity, duration and resolution. Receiving a task identifier does not yet mean the video is ready.
Generated videos last 1 to 15 seconds. 1:1, 16:9, 9:16, 4:3, 3:4, 3:2 or 2:3 can be selected; the standard format for text is 16:9. Image animation retains the image's proportions by default.
480p, 720p and 1080p are available for generation from text and a starting image. 480p is the default; reference mode is limited to 720p. The existence of 1080p in the model does not mean it applies to every mode.
In reference mode, xAI accepts up to 7 images. They define people, objects, clothing and other visual characteristics without fixing the initial frame. This method is intended for scenes with a recurring character, product demonstrations and retention of the desired appearance across different stagings.
Version 1.5 allows the exact first frame, exact last frame or both to be fixed. If only the last is specified, the model constructs the beginning itself and arrives at the supplied image. With both boundaries specified, it creates movement between them; a text instruction can refine motion and camera work.
For storyboarding within a video, up to 4 keyframes can be fixed. Each time must be strictly after the start and before the end; placement follows a 1/3-second grid. Two frames that round to the same grid point are rejected. To avoid this, leave at least 1/3 second between them. First and last frames are specified separately from these four intermediate frames.
Anchor frames can be combined with visual references and preset voices. If at least one frame is fixed, a text description is optional: the image sequence itself defines the scene. Keyframes cannot be used with video-editing mode. These xAI capabilities do not mean they are available in Ailoot's published text-based generation.
A generated video contains an audio track by default; a silent result is also provided. In xAI's voice-reference mode, up to 3 preset voices can be selected and associated with speaking participants in the scene.
Supplying a custom audio file as a voice sample is not a general capability of this model: xAI provides it separately on request. A preset voice and uploaded audio are different types of input.
Text-to-video generation creates a new scene; a starting image defines its beginning; references guide appearance; fixed frames specify visual points in time. Each mode has its own limitations, so settings from one method cannot be unconditionally applied to another.
POST /v1/videos/generationstextvideo| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits grok-imagine-video-1.5-preview | Default— |
Fieldprompt | Typestring | RequirementOptional | Values & limits Minimum length: 1 | Default— |
Fieldduration | Typeinteger | RequirementOptional | Values & limits Minimum: 1 Maximum: 15 Conditions: Do not supply bothdurationandseconds Unit: seconds | Default1 |
Fieldseconds | Typeinteger|integer_string | RequirementOptional | Values & limits Minimum: 1 Maximum: 15 Alternative: duration | Default— |
Fieldresolution | Typestring | RequirementOptional | Values & limits 480p, 720p, 1080p | Default480p |
Fieldaspect_ratio | Typestring | RequirementOptional | Values & limits 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3 Conditions: Ifimageinputsuppliedandratioomitted,providerinheritsimageaspect | Default16:9 |
Fieldsize | Typestring | RequirementOptional | Values & limits 1280x720, 720x1280, 1920x1080, 1080x1920 Conditions: Mustagreewithresolution/aspect_ratioifbothspecified | Default— |
Fieldimage | Typeobject | RequirementOptional | Values & limits Conditions: Actualsupportedinputimage; permitsomittingprompt | Default— |
Text to video · Required: prompt
Image to video · Required: image
| Parameter | Contract |
|---|---|
Parametermaterial_types | Contracttext |
Parameterrequired | Contractnull |
Parameteroptional_controls | Contractnull |
Lifecycle: submit_once_poll_original
GET /v1/videos/{request_id}
GET /v1/videos/generations/{request_id}
GET /v1/videos/{request_id}/content
GET /v1/videos/generations/{request_id}/content
Method shapes and IDs are exact; supplier capability is separate from admission.
POST /v1/videos/generationsAiloot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\ncurl https://api.ailoot.ai/v1/videos/generations \\\n -H "Authorization: Bearer $AILOOT_API_KEY" \\\n -H "Content-Type: application/json" \\\n -d '{"model": "grok-imagine-video-1.5-preview", "prompt": "A person walks slowly through a quiet garden as the camera tracks from the side."}'mp4
GET /v1/videos/{request_id}GET /v1/videos/generations/{request_id}GET /v1/videos/{request_id}/contentGET /v1/videos/generations/{request_id}/contentset -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_VIDEO_ID:?Set the actual Ailoot operation ID returned by creation}"\nexport AILOOT_VIDEO_ID\nvideo_id=$(python3 -c 'import os, urllib.parse; print(urllib.parse.quote(os.environ["AILOOT_VIDEO_ID"], safe=""))') || exit 1\ncurl "https://api.ailoot.ai/v1/videos/generations/$video_id" \\\n -H "Authorization: Bearer $AILOOT_API_KEY"set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_VIDEO_ID:?Set the actual Ailoot operation ID returned by creation}"\nexport AILOOT_VIDEO_ID\nvideo_id=$(python3 -c 'import os, urllib.parse; print(urllib.parse.quote(os.environ["AILOOT_VIDEO_ID"], safe=""))') || exit 1\ncurl "https://api.ailoot.ai/v1/videos/generations/$video_id/content" \\\n -H "Authorization: Bearer $AILOOT_API_KEY" \\\n --output video.mp4Save every completed video or returned file; retrieve every actual returned asset URL.
Keep the complete status JSON and the actual request_id. A completed response with status=done contains video.url with an owned content URL. Preserve the reported duration and every returned result field.
Read existing original job; never resubmit unknown send.
| Field | Type |
|---|---|
Fieldrequest_id | Typestring |
Fieldid | Typestring |
Fieldstatus | Typestring |
Fieldvideo.url | Typeowned signed content URL |
Fieldvideo.duration | Typenumber |
Fieldvideo.respect_moderation | Typeboolean |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Response409 · video_not_ready | ConditionOriginal result content is not yet available | Next stepPoll the original returned ID and respect Retry-After. |
Responseserver_error | ConditionThe returned response says the operation or result requires a check | Next stepKeep the original returned operation/task ID, request headers and response; continue supported reads of that original operation. Do not resubmit the creation request. |
Ailoot grok-imagine-video-1.5-preview
Use the method paths shown above.