logo STATICALIZA
Home Projects Socials API
Daily Credits Usage 0 / 1000 used

Resets at --:-- --

INFRASTRUCTURE

Browse Stativerse AI model endpoints by infrastructure.

V1

Shared endpoints running on slower, shared hosted models.

ENDPOINTS

Authorization is required on every request. Use your Stativerse API key in the Authorization: Bearer <key> header. You can create an API key in your account settings. Click any endpoint path to copy the full endpoint shape.

Select a model card to view its endpoint configuration.

POST /models/v1/image-generation >

Payload

prompt string Text prompt used to generate the image.
negative_prompt string Optional negative prompt for undesired elements.
resolution [...] Optional generation size. Defaults to [1024, 1024]. >
[0] number Width in pixels.
[1] number Height in pixels.
post_resolution [...] Optional final output size after crop or resize. Defaults to resolution. >
[0] number Width in pixels.
[1] number Height in pixels.
steps number Optional diffusion step count. Defaults to 8.
guidance number Optional true CFG guidance scale. Defaults to 1.
crop boolean If true, centers and crops to aspect ratio instead of resizing.
remove_background boolean If true, attempts to remove background and return RGBA format.
safety_check boolean Runs text and output image safety classifiers. Also enabled when safety_input is provided.
safety_input string Optional instruction prepended to the prompt before text safety classification.
output_format string Allowed values: "Base64 PNG", "Base64 JPEG", "Base64 JPEG Progressive", "RGBA 2D", "RGBA 1D". Defaults to "Base64 PNG".
seed number Optional random seed. Defaults to a generated seed.

Response

ok boolean True when the request succeeds. False when it fails.
data {...} Response payload when ok is true. >
id string Unique task generation id from the upstream model backend.
output string | number[] | number[][][] Encoded image data or RGBA array output based on output_format.
seed number Random seed used by generator in run.
generation_time number Server-side generation duration in seconds.
nsfw_input_classifier {...} | null Present when safety_check is enabled. >
output [...] Per-label prompt safety scores. >
[...] {...} Object shape for each scored label entry. >
label string Safety label name.
score number Score for the current label.
highest {...} Highest-scoring safety label object. >
label string Safety label name.
score number Highest returned score.
lowest {...} Lowest-scoring safety label object. >
label string Safety label name.
score number Lowest returned score.
nsfw_output_classifier string | null Visual safety label returned for the generated image when safety_check is enabled.
error {...} Null when ok is true. On failure includes public error details. >
code string Normalized public error code.
message string Detailed error mapping or helper error format.
POST /models/v1/image-edit >

Payload

prompt string Optional edit instructions. If omitted, the input image is returned after processing.
negative_prompt string Optional negative prompt for details you want the edit to avoid.
image base64 | [...] Required base64 image or image array. Data URI prefixes are accepted. >
[...] base64 Base64 image item.
resolution [...] Optional edit generation size. Omit to preserve input size. >
[0] number Width in pixels.
[1] number Height in pixels.
post_resolution [...] Optional final output size after crop or resize. Defaults to resolution or input size. >
[0] number Width in pixels.
[1] number Height in pixels.
steps number Optional diffusion step count. Defaults to 8.
guidance number Optional true CFG guidance scale. Defaults to 1.
crop boolean If true, centers and crops to aspect ratio instead of resizing.
remove_background boolean If true, attempts to remove background and return RGBA format.
safety_check boolean Runs text and output image safety classifiers. Also enabled when safety_input is provided.
safety_input string Optional instruction prepended to the prompt before text safety classification.
output_format string Allowed values: "Base64 PNG", "Base64 JPEG", "Base64 JPEG Progressive", "RGBA 2D", "RGBA 1D". Defaults to "Base64 PNG".
seed number Optional random seed. Defaults to a generated seed.

Response

ok boolean True when the request succeeds. False when it fails.
data {...} Response payload when ok is true. >
id string Unique task generation id from the upstream model backend.
output string | number[] | number[][][] Encoded image data or RGBA array output based on output_format.
seed number Random seed used by generator in run.
generation_time number Server-side generation duration in seconds.
nsfw_input_classifier {...} | null Present when safety_check is enabled. >
output [...] Per-label prompt safety scores. >
[...] {...} Object shape for each scored label entry. >
label string Safety label name.
score number Score for the current label.
highest {...} Highest-scoring safety label object. >
label string Safety label name.
score number Highest returned score.
lowest {...} Lowest-scoring safety label object. >
label string Safety label name.
score number Lowest returned score.
nsfw_output_classifier string | null Visual safety label returned for the edited image when safety_check is enabled.
error {...} Null when ok is true. On failure includes public error details. >
code string Normalized public error code.
message string Detailed error mapping or helper error format.
POST /models/v1/classification >

Payload

input string | [...] Required text or text array to classify. >
[...] string Batch input text item.
choices [...] Required labels array of categories to score against the input. >
[...] string Candidate classification label.
normalize boolean Use exp distribution probability array normalizations over absolute limits. Defaults to True.

Response

ok boolean True when the request succeeds. False when it fails.
data {...} Response payload when ok is true. >
id string Unique task generation id from the upstream model backend.
output [...] | [[...], ...] Sorted classification label scores. Batch input returns an array of result arrays. >
label string Candidate label value from choices.
score number Classification score for the current label.
highest {...} | [...] Highest-scoring item for a single input, or one highest item per batch entry. >
label string Highest-scoring classification label.
score number Highest classification score.
lowest {...} | [...] Lowest-scoring item for a single input, or one lowest item per batch entry. >
label string Lowest-scoring classification label.
score number Lowest classification score.
generation_time number Server-side execution duration in seconds.
error {...} Null when ok is true. On failure includes public error details. >
code string Normalized public error code.
message string Detailed error mapping or helper error format.
POST /models/v1/sentence >

Payload

input string | [...] Required source sentence, query, or string array to compare. >
[...] string Batch source text item.
choices [...] Required candidate sentence strings to score against the input. >
[...] string Candidate sentence string.
normalize boolean Use exp distribution probability array normalizations over absolute limits. Defaults to True.

Response

ok boolean True when the request succeeds. False when it fails.
data {...} Response payload when ok is true. >
id string Unique task generation id from the upstream model backend.
output [...] | [[...], ...] Sorted sentence-similarity scores. Batch input returns an array of result arrays. >
label string Candidate sentence value from choices.
score number Sentence-similarity score for the current choice.
highest {...} | [...] Highest-scoring item for a single input, or one highest item per batch entry. >
label string Highest-scoring candidate sentence.
score number Highest sentence-similarity score.
lowest {...} | [...] Lowest-scoring item for a single input, or one lowest item per batch entry. >
label string Lowest-scoring candidate sentence.
score number Lowest sentence-similarity score.
generation_time number Server-side execution duration in seconds.
error {...} Null when ok is true. On failure includes public error details. >
code string Normalized public error code.
message string Detailed error mapping or helper error format.
© Staticaliza