Stable Diffusion

Generate images from text prompts

post

Generate images from text using Stable Diffusion (Automatic1111).

Available Models: Realistic_Vision_V5.1.safetensors, illustriousRealismBy_v10VAE.safetensors, stable-diffusion-v1-5-pruned.safetensors

Required Parameters

Parameter	Type	Description
prompt	string	Image generation prompt
model	string	SD checkpoint (see available models above)

Optional Parameters - Basic

Parameter	Type	Default	Description
negative_prompt	string	""	Things to avoid in the image
width	integer	832	Image width (64-2048)
height	integer	1216	Image height (64-2048)
steps	integer	30	Sampling steps (1-150)
cfg_scale	float	4.5	Prompt adherence (1.0-30.0)
seed	integer	-1	Random seed (-1 = random)
batch_size	integer	1	Images per batch (1-8)
sampler_name	string	DPM++ SDE	Sampler algorithm
scheduler	string	Karras	Noise scheduler

Optional Parameters - Hires Fix

Parameter	Type	Default	Description
enable_hr	boolean	true	Enable hires upscaling
hr_scale	float	1.5	Upscale ratio (1.0-4.0)
hr_upscaler	string	Latent	Upscaler type
hr_second_pass_steps	integer	15	Hires sampling steps
denoising_strength	float	0.45	Hires denoising (0.0-1.0)
hr_sampler_name	string	-	Sampler for hires pass
hr_scheduler	string	-	Scheduler for hires pass
hr_resize_x	integer	-	Target width (0 = use hr_scale)
hr_resize_y	integer	-	Target height (0 = use hr_scale)
hr_prompt	string	-	Prompt for hires pass
hr_negative_prompt	string	-	Negative prompt for hires pass

Optional Parameters - Advanced

Parameter	Type	Default	Description
n_iter	integer	-	Number of batches
subseed	integer	-	Subseed for variations
subseed_strength	float	-	Variation strength (0.0-1.0)
restore_faces	boolean	-	Enable face restoration
tiling	boolean	-	Generate tileable images
styles	array	-	Style presets to apply

Async Mode

Set async: true to get a task_address immediately and poll for results.

Async Response (202):

{
  "task_address": "direct:abc12345-...",
  "status": "running",
  "poll_url": "/v2/tasks/direct:abc12345-..."
}

Poll GET /v2/tasks/{task_address} until status: completed.

Header parameters

x-api-keyany ofOptional

stringOptional

or

nullOptional

Body

application/json

Request schema for Stable Diffusion txt2img generation.

Compatible with Automatic1111 API format.

promptstring · min: 1Required

Image generation prompt (supports LoRA tags)

Example: masterpiece, best quality, a serene mountain landscape at sunset, dramatic lighting

negative_promptany ofOptional

Negative prompt (things to avoid)

Default: ""Example: low quality, blurry, watermark, text

stringOptional

or

nullOptional

modelstringRequired

SD checkpoint. Available: illustriousRealismBy_v10VAE.safetensors, Realistic_Vision_V5.1.safetensors, stable-diffusion-v1-5-pruned.safetensors

Example: Realistic_Vision_V5.1.safetensors

widthany ofOptional

Image width in pixels

Default: 832

integer · min: 64 · max: 2048Optional

or

nullOptional

heightany ofOptional

Image height in pixels

Default: 1216

integer · min: 64 · max: 2048Optional

or

nullOptional

stepsany ofOptional

Sampling steps (more = higher quality, slower)

Default: 30

integer · min: 1 · max: 150Optional

or

nullOptional

sampler_nameany ofOptional

Sampler algorithm

Default: DPM++ SDE

stringOptional

or

nullOptional

schedulerany ofOptional

Noise scheduler

Default: Karras

stringOptional

or

nullOptional

cfg_scaleany ofOptional

CFG scale (prompt adherence, higher = stricter)

Default: 4.5

number · min: 1 · max: 30Optional

or

nullOptional

denoising_strengthany ofOptional

Denoising strength for hires fix

Default: 0.45

number · max: 1Optional

or

nullOptional

enable_hrany ofOptional

Enable hires fix (upscaling)

Default: true

booleanOptional

or

nullOptional

hr_scaleany ofOptional

Hires upscale ratio

Default: 1.5

number · min: 1 · max: 4Optional

or

nullOptional

hr_upscalerany ofOptional

Hires upscaler type

Default: Latent

stringOptional

or

nullOptional

hr_second_pass_stepsany ofOptional

Hires second pass steps

Default: 15

integer · max: 150Optional

or

nullOptional

seedany ofOptional

Random seed (-1 = random)

Default: -1

integerOptional

or

nullOptional

batch_sizeany ofOptional

Number of images to generate

Default: 1

integer · min: 1 · max: 8Optional

or

nullOptional

subseedany ofOptional

Subseed for variations (-1 = random)

integerOptional

or

nullOptional

subseed_strengthany ofOptional

Subseed strength (0 = no variation, 1 = full variation)

number · max: 1Optional

or

nullOptional

n_iterany ofOptional

Number of batches to generate

integer · min: 1 · max: 10Optional

or

nullOptional

stylesany ofOptional

Style presets to apply

string[]Optional

or

nullOptional

restore_facesany ofOptional

Enable face restoration

booleanOptional

or

nullOptional

tilingany ofOptional

Generate tileable/seamless images

booleanOptional

or

nullOptional

hr_resize_xany ofOptional

Hires fix target width (0 = use hr_scale)

integer · max: 4096Optional

or

nullOptional

hr_resize_yany ofOptional

Hires fix target height (0 = use hr_scale)

integer · max: 4096Optional

or

nullOptional

hr_sampler_nameany ofOptional

Sampler for hires fix pass (default: same as sampler_name)

stringOptional

or

nullOptional

hr_schedulerany ofOptional

Scheduler for hires fix pass (default: same as scheduler)

stringOptional

or

nullOptional

hr_promptany ofOptional

Prompt for hires fix pass (default: same as prompt)

stringOptional

or

nullOptional

hr_negative_promptany ofOptional

Negative prompt for hires fix pass (default: same as negative_prompt)

stringOptional

or

nullOptional

modeany ofOptional

Routing mode: 'auto' (intelligent), 'opengpu' (blockchain), 'direct' (low-latency)

Default: auto

string · enumOptionalPossible values:

or

nullOptional

asyncany ofOptional

Return task_address immediately, poll /v2/tasks/{task_address} for result. Default: false (synchronous).

Default: false

booleanOptional

or

nullOptional

Responses

200

Successful Response

application/json

202

Task accepted (async mode). Poll the poll_url for status.

application/json

422

Validation Error

application/json

post

/v2/automatic1111/sdapi/v1/txt2img

HTTPcURLJavaScriptPython

POST /v2/automatic1111/sdapi/v1/txt2img HTTP/1.1
Content-Type: application/json
Accept: */*
Content-Length: 90

{
  "prompt": "a cute fluffy cat, photorealistic",
  "model": "Realistic_Vision_V5.1.safetensors"
}

Minimal (Recommended)With Quality SettingsWith Hires FixAll Options

{
  "generated_at": "2025-12-16T10:30:00Z",
  "images": [
    "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mNk+M9QDwADhgGAWjR9awAAAABJRU5ErkJggg=="
  ],
  "info": {
    "cfg_scale": 4.5,
    "height": 1216,
    "sampler_name": "DPM++ SDE",
    "seed": 42,
    "steps": 30,
    "width": 832
  }
}