Higgsfield
App · 16 actions
Make AI images and videos with Higgsfield's models (SOUL, Kling, Seedance, Wan and more), follow and tidy up generations, and train Soul ID characters.
Actions
| Action | What it does |
|---|---|
| Cancel a generation | Cancel a generation that is still waiting to start, and get the money back. |
| Create a character | Train a Soul ID character from photos of one person, to feature them in SOUL images. |
| Delete a generation | Delete a finished generation's files, prompt and settings from Higgsfield now, instead of after 7 days. |
| Estimate cost | See what a generation would cost, in US dollars, without running it or being charged. |
| Generate a video from an image | Animate an image into a video, optionally ending on a second image. |
| Generate a video from references | Make a video guided by reference images, videos or audio, such as a product or a person to feature. |
| Generate a video from text | Make a video from a prompt with a model such as Kling, Seedance, Wan or MiniMax. |
| Generate an image | Make images from a prompt, or edit reference images, with a Higgsfield image model such as SOUL V2, Marketing Studio or Recraft. |
| Get a character | Get a Soul ID character and whether its training has finished. |
| Get a generation | Check a generation's progress, and get its images or video once it has finished. |
| List Marketing Studio presets | List the ready-made ad looks Marketing Studio Image can use, optionally searching by name. |
| List Restyle styles | List the looks Genjutsu Restyle can give a video. |
| List SOUL styles | List the styles SOUL or SOUL V2 can give an image. |
| Run any Higgsfield model | Run any Higgsfield model, such as Cinema Studio, by its endpoint ID with its own settings. |
| Transform a video | Edit or extend a video, restyle it, swap an object in it, or move a character the way the video moves. |
| Wait for a generation | Wait until a generation finishes, then return it. |
Cancel a generation
Cancel a generation that is still waiting to start, and get the money back. One that has already started can't be canceled: the step says so, with its status, instead of failing.
Inputs
- Request ID
request_idtextRequired - The generation's request ID, from a generation step.
Outputs
canceledtrue/false- True if it was canceled; false if it had already started
request_idtext- The request
statustext- canceled, or the status it had when it couldn't be (in_progress, completed, failed or nsfw)
Example output
{
"canceled": true,
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "canceled"
}Create a character
Train a Soul ID character from photos of one person, to feature them in SOUL images. Training takes a while: check it with 'Get a character'.
Inputs
- Name
nametextRequired - A name for the character, up to 100 characters.
- Photos
image_urlslistRequired - Public addresses of 1 to 100 photos of the person, one per entry. Several angles and settings work best.
- For model
model_versiontext - The SOUL model that will use the character. Defaults to SOUL.One of:
v1(SOUL (default)),v2(SOUL V2),cinema(SOUL Cinema). Default:v1.
Outputs
characterobject- The Soul ID character
Example output
{
"character": {
"created_at": "2026-11-06T12:00:00Z",
"fail_reason": "",
"id": "5a1c2f0e-0f3b-4c7e-9a52-6a4c1d2b3e4f",
"model_version": "v2",
"name": "Jo",
"ready": false,
"status": "queued",
"thumbnail_url": ""
}
}Delete a generation
Delete a finished generation's files, prompt and settings from Higgsfield now, instead of after 7 days.
Inputs
- Request ID
request_idtextRequired - The generation's request ID, from a generation step.
Outputs
deletedtrue/false- Always true once deleted
request_idtext- The request deleted
Example output
{
"deleted": true,
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff"
}Estimate cost
See what a generation would cost, in US dollars, without running it or being charged.
Inputs
- Model
modeltextRequired - The model's endpoint ID, such as higgsfield-ai/soul/v2/standard. Pick one, or type the ID shown on any model's page at docs.higgsfield.ai.
- Settings
parametersobject - The model's settings, by the names on its page at docs.higgsfield.ai, such as {"prompt": "A red fox in snow", "aspect_ratio": "16:9"}.
Outputs
modeltext- The model's endpoint ID
usdtext- The cost in US dollars, as text
usd_amountnumber- The cost in US dollars
Example output
{
"model": "higgsfield-ai/soul/v2/standard",
"usd": "0.094",
"usd_amount": 0.094
}Generate a video from an image
Animate an image into a video, optionally ending on a second image. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The Higgsfield model that makes the video.One of:
minimax/hailuo-2.3/standard/image-to-video(Hailuo 2.3: Standard image to video),alibaba/happy-horse/image-to-video(Happy Horse 1.0: Image to video),alibaba/happy-horse/v1.1/image-to-video(HappyHorse 1.1: Image to video),kling-video/v2.5-turbo/pro/image-to-video(Kling 2.5 Turbo: Pro image to video),kling-video/v2.5-turbo/standard/image-to-video(Kling 2.5 Turbo: Standard image to video),kling-video/v2.6/pro/image-to-video(Kling 2.6: Pro image to video),kling-video/v3.0/4k/image-to-video(Kling 3.0: 4K, Image to video),kling-video/v3.0/pro/image-to-video(Kling 3.0: Pro, Image to video),kling-video/v3.0/std/image-to-video(Kling 3.0: Standard, Image to video),kling-video/v3.0-turbo/image-to-video(Kling 3.0: Turbo, Image to video), and 10 more. - Image
image_urltextRequired - The public address of the image to animate (the first frame).
- Prompt
prompttext - Describe what should happen. Some models need one.
- Last frame
end_image_urltext - The public address of an image the video should end on. Only some models take this.
- Length (seconds)
durationnumber - How long the video is. Each model has its own lengths (such as 5 or 10); the step says which if this one doesn't fit.
- Aspect ratio
aspect_ratiotext - The video's shape, width to height. Leave empty for the model's default.One of:
auto(Auto (the model decides)),adaptive(Match the input),21:9,16:9,3:2,4:3,1:1,3:4,2:3,9:16. - Resolution
resolutiontext - The video's resolution. Higher costs more. Leave empty for the model's default.One of:
360p,480p,540p,720p,1080p,2k,4k. - Sound
soundtext - Whether the video gets generated sound. Leave empty for the model's default; models without sound don't take it.One of:
on,off. - Improve the prompt
enhance_prompttext - Let the model expand the prompt before making the result. Leave empty for the model's default.One of:
on,off. - Leave out
negative_prompttext - What the result should not show. Only some models take this.
- Seed
seednumber - A number that makes the result repeatable: the same seed and settings give a similar result. Leave empty for a random one.
- More settings
more_settingsobject - Any other setting the chosen model takes, by the name on its page at docs.higgsfield.ai, such as cfg_scale or camera_movement.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "",
"images": [],
"model": "kling-video/v2.5-turbo/pro/image-to-video",
"output_urls": [
"https://cdn.higgsfield.ai/example/video.mp4"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": "https://cdn.higgsfield.ai/example/video.mp4"
}Generate a video from references
Make a video guided by reference images, videos or audio, such as a product or a person to feature. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The Higgsfield model that makes the video.One of:
xai/grok-imagine-video/v1.5/reference-to-video(Grok Imagine Video 1.5: Reference to video),alibaba/happy-horse/reference-to-video(Happy Horse 1.0: Reference to video),alibaba/happy-horse/v1.1/reference-to-video(HappyHorse 1.1: Reference to video),kling-video/o3/image-reference(Kling O3: Image reference),kling-video/o3/video-reference(Kling O3: Video reference),kling-video/omni/image-reference(Kling Omni: Image reference),kling-video/omni/video-reference(Kling Omni: Video reference),minimax/h3/reference-to-video(MiniMax H3: Reference to video),bytedance/seedance-2.0/reference-to-video(Seedance 2.0: Reference to video),bytedance/seedance-2.5/reference-to-video(Seedance 2.5: Reference to video), and 4 more. - Prompt
prompttext - Describe what should happen. Some models need one.
- Reference images
reference_image_urlslist - Public addresses of images to guide the result, one per entry. Models that take a single image use the first.
- Reference videos
reference_video_urlslist - Public addresses of videos to guide the result, one per entry.
- Reference audio
reference_audio_urlslist - Public addresses of audio files to guide the result, one per entry.
- Length (seconds)
durationnumber - How long the video is. Each model has its own lengths (such as 5 or 10); the step says which if this one doesn't fit.
- Aspect ratio
aspect_ratiotext - The video's shape, width to height. Leave empty for the model's default.One of:
auto(Auto (the model decides)),adaptive(Match the input),21:9,16:9,3:2,4:3,1:1,3:4,2:3,9:16. - Resolution
resolutiontext - The video's resolution. Higher costs more. Leave empty for the model's default.One of:
360p,480p,540p,720p,1080p,2k,4k. - Sound
soundtext - Whether the video gets generated sound. Leave empty for the model's default; models without sound don't take it.One of:
on,off. - Seed
seednumber - A number that makes the result repeatable: the same seed and settings give a similar result. Leave empty for a random one.
- More settings
more_settingsobject - Any other setting the chosen model takes, by the name on its page at docs.higgsfield.ai, such as cfg_scale or camera_movement.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "",
"images": [],
"model": "bytedance/seedance-2.0/reference-to-video",
"output_urls": [
"https://cdn.higgsfield.ai/example/video.mp4"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": "https://cdn.higgsfield.ai/example/video.mp4"
}Generate a video from text
Make a video from a prompt with a model such as Kling, Seedance, Wan or MiniMax. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The Higgsfield model that makes the video.One of:
minimax/hailuo-2.3/standard/text-to-video(Hailuo 2.3: Standard text to video),alibaba/happy-horse/text-to-video(Happy Horse 1.0: Text to video),alibaba/happy-horse/v1.1/text-to-video(HappyHorse 1.1: Text to video),kling-video/v2.5-turbo/pro/text-to-video(Kling 2.5 Turbo: Pro text to video),kling-video/v2.6/pro/text-to-video(Kling 2.6: Pro text to video),kling-video/v3.0/4k/text-to-video(Kling 3.0: 4K, Text to video),kling-video/v3.0/pro/text-to-video(Kling 3.0: Pro, Text to video),kling-video/v3.0/std/text-to-video(Kling 3.0: Standard, Text to video),kling-video/v3.0-turbo/text-to-video(Kling 3.0: Turbo, Text to video),lightricks/ltx-2.5/text-to-video/fast(LTX-2.5: Text to video fast), and 9 more. - Prompt
prompttextRequired - Describe what to make.
- Length (seconds)
durationnumber - How long the video is. Each model has its own lengths (such as 5 or 10); the step says which if this one doesn't fit.
- Aspect ratio
aspect_ratiotext - The video's shape, width to height. Leave empty for the model's default.One of:
auto(Auto (the model decides)),adaptive(Match the input),21:9,16:9,3:2,4:3,1:1,3:4,2:3,9:16. - Resolution
resolutiontext - The video's resolution. Higher costs more. Leave empty for the model's default.One of:
360p,480p,540p,720p,1080p,2k,4k. - Sound
soundtext - Whether the video gets generated sound. Leave empty for the model's default; models without sound don't take it.One of:
on,off. - Improve the prompt
enhance_prompttext - Let the model expand the prompt before making the result. Leave empty for the model's default.One of:
on,off. - Leave out
negative_prompttext - What the result should not show. Only some models take this.
- Seed
seednumber - A number that makes the result repeatable: the same seed and settings give a similar result. Leave empty for a random one.
- More settings
more_settingsobject - Any other setting the chosen model takes, by the name on its page at docs.higgsfield.ai, such as cfg_scale or camera_movement.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "",
"images": [],
"model": "bytedance/seedance-2.0/text-to-video",
"output_urls": [
"https://cdn.higgsfield.ai/example/video.mp4"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": "https://cdn.higgsfield.ai/example/video.mp4"
}Generate an image
Make images from a prompt, or edit reference images, with a Higgsfield image model such as SOUL V2, Marketing Studio or Recraft. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The Higgsfield model that makes the images.One of:
higgsfield/ads-studio/v1.0(Ads Studio: Generate ads),xai/grok-imagine-image-2.0(Grok Image 2.0: Generate and edit),ideogram/v4.0(Ideogram 4.0: Generate),marketing-studio/image(Marketing Studio Image: 2.0 Alpha),marketing-studio/image/flare(Marketing Studio Image: 2.5 Flare),marketing-studio/image/sunburst(Marketing Studio Image: 2.5 Sunburst),alibaba/qwen-image-3/edit(Qwen Image 3: Edit images),alibaba/qwen-image-3/text-to-image(Qwen Image 3: Text to image),recraft/v4.1/text-to-image(Recraft V4.1: Text to image),recraft/v4.1/pro/text-to-image(Recraft V4.1 Pro: Text to image), and 6 more. - Prompt
prompttextRequired - Describe what to make.
- Aspect ratio
aspect_ratiotext - The image's shape, width to height. Leave empty for the model's default.One of:
auto(Auto (the model decides)),3:1,8:3,23:9,22:9,12:5,21:9,2:1,16:9,8:5, and 20 more. - Resolution
resolutiontext - The image's resolution. Leave empty for the model's default.One of:
720p,1080p,1k,2k,4k. - Reference images
reference_image_urlslist - Public addresses of images to guide the result, one per entry. Models that take a single image use the first.
- Number of images
number_of_imagesnumber - How many images to make. SOUL models make 1 or 4; Ads Studio 1 to 8. Each one is charged.From 1 to 8.
- Quality
qualitytext - For Marketing Studio and Grok Image: higher looks better and costs more. Leave empty for the model's default.One of:
low,medium,high,xhigh(Extra high),max(Maximum). - Style
style_idtext - For SOUL and SOUL V2: the look of the image. Pick one, or give its ID.
- Character
character_idtext - For SOUL models: a trained Soul ID character's ID (from 'Create a character'), to put that person in the image.
- Preset
preset_idtext - For Marketing Studio: a ready-made ad look. Pick one, or give its ID.
- Improve the prompt
enhance_prompttext - Let the model expand the prompt before making the result. Leave empty for the model's default.One of:
on,off. - Leave out
negative_prompttext - What the result should not show. Only some models take this.
- Seed
seednumber - A number that makes the result repeatable: the same seed and settings give a similar result. Leave empty for a random one.
- More settings
more_settingsobject - Any other setting the chosen model takes, by the name on its page at docs.higgsfield.ai, such as cfg_scale or camera_movement.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "https://cdn.higgsfield.ai/example/image-1.jpg",
"images": [
{
"url": "https://cdn.higgsfield.ai/example/image-1.jpg"
}
],
"model": "higgsfield-ai/soul/v2/standard",
"output_urls": [
"https://cdn.higgsfield.ai/example/image-1.jpg"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": ""
}Get a character
Get a Soul ID character and whether its training has finished.
Inputs
- Character ID
character_idtextRequired - The character's ID, from 'Create a character'.
Outputs
characterobject- The Soul ID character
Example output
{
"character": {
"created_at": "2026-11-06T12:00:00Z",
"fail_reason": "",
"id": "5a1c2f0e-0f3b-4c7e-9a52-6a4c1d2b3e4f",
"model_version": "v2",
"name": "Jo",
"ready": true,
"status": "completed",
"thumbnail_url": ""
}
}Get a generation
Check a generation's progress, and get its images or video once it has finished.
Inputs
- Request ID
request_idtextRequired - The generation's request ID, from a generation step.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "https://cdn.higgsfield.ai/example/image-1.jpg",
"images": [
{
"url": "https://cdn.higgsfield.ai/example/image-1.jpg"
}
],
"output_urls": [
"https://cdn.higgsfield.ai/example/image-1.jpg"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": ""
}List Marketing Studio presets
List the ready-made ad looks Marketing Studio Image can use, optionally searching by name.
Inputs
- Search
searchtext - Only presets or groups whose name contains this.
- Page size
page_sizenumber - How many to return, 1 to 100. Defaults to 50.From 1 to 100.
- Next page cursor
cursornumber - The 'next_cursor' from an earlier page, to get the page after it.
Outputs
countnumber- How many on this page
has_moretrue/false- True if there are more
next_cursornumber- The cursor for the next page, empty on the last
presetslist- The presets
totalnumber- How many there are in all
Example output
{
"count": 1,
"has_more": true,
"next_cursor": 50,
"presets": [
{
"aspect_ratio": "3:4",
"cover_image": "",
"group_name": "Editorial",
"id": "8b0c5e1a-2d3f-4a6b-9c7d-1e2f3a4b5c6d",
"name": "Editorial product portrait",
"type": "ads"
}
],
"total": 120
}List Restyle styles
List the looks Genjutsu Restyle can give a video.
Inputs
None.
Outputs
countnumber- How many
styleslist- The styles
Example output
{
"count": 1,
"styles": [
{
"description": "",
"id": "c2143317-f28d-4c3c-a0b8-39bd547e08a7",
"name": "Cel-Shaded CG Anime",
"preview_url": "https://cdn.higgsfield.ai/example/restyle.webp"
}
]
}List SOUL styles
List the styles SOUL or SOUL V2 can give an image.
Inputs
- Model
versiontext - Whose styles to list. Defaults to SOUL V2.One of:
v2(SOUL V2 (default)),v1(SOUL). Default:v2.
Outputs
countnumber- How many
styleslist- The styles
versiontext- v2 or v1
Example output
{
"count": 1,
"styles": [
{
"description": "A polished, all-round style.",
"id": "3db34ab5-3439-4317-9e03-08dc30852e69",
"name": "General",
"preview_url": "https://cdn.higgsfield.ai/example/style.jpg"
}
],
"version": "v2"
}Run any Higgsfield model
Run any Higgsfield model, such as Cinema Studio, by its endpoint ID with its own settings. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The model's endpoint ID, such as higgsfield-ai/soul/v2/standard. Pick one, or type the ID shown on any model's page at docs.higgsfield.ai.
- Settings
parametersobject - The model's settings, by the names on its page at docs.higgsfield.ai, such as {"prompt": "A red fox in snow", "aspect_ratio": "16:9"}.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "",
"images": [],
"model": "higgsfield/cinema-studio/4.0",
"output_urls": [
"https://cdn.higgsfield.ai/example/video.mp4"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": "https://cdn.higgsfield.ai/example/video.mp4"
}Transform a video
Edit or extend a video, restyle it, swap an object in it, or move a character the way the video moves. Charged to your Higgsfield balance.
Inputs
- Model
modeltextRequired - The Higgsfield model that makes the new video.One of:
higgsfield/genjutsu/motion-transfer/v1.0(Genjutsu: Motion transfer),higgsfield/genjutsu/object-swap/v1.0(Genjutsu: Object swap),higgsfield/genjutsu/restyle/v1.0(Genjutsu: Restyle),kling-video/motion-control/pro(Kling 2.6 Motion Control: Pro),kling-video/motion-control/std(Kling 2.6 Motion Control: Standard),kling-video/v3/motion-control/pro(Kling 3.0 Motion Control: Pro),kling-video/v3/motion-control/std(Kling 3.0 Motion Control: Standard),kling-video/o3/video-edit(Kling O3: Video edit),kling-video/omni/video-edit(Kling Omni: Video edit),bytedance/seedance-2.5/video-edit(Seedance 2.5: Video edit),bytedance/seedance-2.5/video-extend(Seedance 2.5: Video extend). - Video
video_urltextRequired - The public address of the video to change (an MP4 file, not a player page). Genjutsu needs at least 4 seconds.
- Prompt
prompttext - Describe the change. Edit and extend models need one.
- Character images
reference_image_urlslist - Public addresses of images of who or what to put in the video, one per entry. Motion Control, motion transfer and object swap need one.
- Restyle style
restyle_style_idtext - For Genjutsu Restyle: the look to give the video. Pick one, or give its ID.
- Add (seconds)
durationnumber - For Seedance 2.5 Video extend: how many seconds to add, 4 to 30.
- Resolution
resolutiontext - The video's resolution. Higher costs more. Leave empty for the model's default.One of:
360p,480p,540p,720p,1080p,2k,4k. - Sound
soundtext - Whether the video gets generated sound. Leave empty for the model's default; models without sound don't take it.One of:
on,off. - More settings
more_settingsobject - Any other setting the chosen model takes, by the name on its page at docs.higgsfield.ai, such as cfg_scale or camera_movement.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10. - Notify this address
webhook_urltext - An https:// address of your own that Higgsfield sends the result to when it finishes. Leave empty if you don't have one.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
modeltext- The model's endpoint ID
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "",
"images": [],
"model": "bytedance/seedance-2.5/video-edit",
"output_urls": [
"https://cdn.higgsfield.ai/example/video.mp4"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": "https://cdn.higgsfield.ai/example/video.mp4"
}Wait for a generation
Wait until a generation finishes, then return it. A failed or blocked one is returned with its status, not as an error.
Inputs
- Request ID
request_idtextRequired - The generation's request ID, from a generation step.
- Wait up to (minutes)
wait_minutesnumber - How long this step waits for the result. If it isn't ready by then, the step succeeds with 'finished' false, and 'Wait for a generation' can pick it up. Set 0 to carry on at once.From 0 to 30. Default:
10.
Outputs
audio_urltext- The audio made, if any
cancel_urltext- Higgsfield's address for canceling this request
errortext- Why it failed, if Higgsfield said
expires_attext- When Higgsfield deletes the files (7 days after it finishes); copy them somewhere before then
finishedtrue/false- True once the request has stopped, however it ended
image_urltext- The first image made, empty for a video
imageslist- Every image made
output_urlslist- Every file made, images, video, audio and any others
request_idtext- The request's ID, for 'Get' or 'Wait for a generation'
statustext- queued, in_progress, completed, failed, nsfw (blocked by moderation) or canceled
status_urltext- Higgsfield's address for this request's status
succeededtrue/false- True when it completed with output
video_urltext- The video made, empty for an image
Example output
{
"audio_url": "",
"cancel_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/cancel",
"error": "",
"expires_at": "2026-11-13T12:05:12.481Z",
"finished": true,
"image_url": "https://cdn.higgsfield.ai/example/image-1.jpg",
"images": [
{
"url": "https://cdn.higgsfield.ai/example/image-1.jpg"
}
],
"output_urls": [
"https://cdn.higgsfield.ai/example/image-1.jpg"
],
"request_id": "d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff",
"status": "completed",
"status_url": "https://api.higgsfield.ai/requests/d7e6c0f3-6699-4f6c-bb45-2ad7fd9158ff/status",
"succeeded": true,
"video_url": ""
}