Stable Diffusion 3 Medium Text to Image Serverless API

Stable Diffusion is a type of latent diffusion model that can generate images from text. It was created by a team of researchers and engineers from CompVis, Stability AI, and LAION. Stable Diffusion v2 is a specific version of the model architecture. It utilizes a downsampling-factor 8 autoencoder with an 865M UNet and OpenCLIP ViT-H/14 text encoder for the diffusion model. When using the SD 2-v model, it produces 768x768 px images. It uses the penultimate text embeddings from a CLIP ViT-H/14 text encoder to condition the generation process.

~7.37s

~$0.041

Playground API

Pricing

~7.37s

~$0.041

 1import requests
 2import json
 3
 4url = "https://api.segmind.com/v1/stable-diffusion-3-medium-txt2img"
 5headers = {
 6    "x-api-key": "YOUR_API_KEY",
 7    "Content-Type": "application/json"
 8}
 9
10data = {
11    "prompt": "A whimsical and high-resolution highly realistic image of a panda in a vintage cosmonaut suit. The panda is holding a sign that reads 'I love flying to the moon!' in playful lettering. The panda's helmet has a small propeller on top and a Indian flag patch, adding to the cosmic vibe. The background features a retro-styled spaceship with rockets and stars, giving the impression of a thrilling journey through space",
12    "negative_prompt": "bad quality, poor quality, doll, disfigured, jpg, toy, bad anatomy, missing limbs, missing fingers, 3d, cgi",
13    "samples": 1,
14    "scheduler": "DPM++ 2M",
15    "num_inference_steps": 25,
16    "guidance_scale": 5,
17    "denoise": 1,
18    "seed": 468685,
19    "img_width": 1024,
20    "img_height": 1024,
21    "modelsamplingsd3_shift": 3,
22    "conditioningsettimesteprange_start": 0.1,
23    "conditioningsettimesteprange_stop": 1,
24    "base64": false
25}
26
27response = requests.post(url, headers=headers, json=data)
28
29if response.status_code == 200:
30    result = response.json()
31    print(json.dumps(result, indent=2))
32else:
33    print(f"Error: {response.status_code}")
34    print(response.text)

 1import requests
 2import json
 3
 4url = "https://api.segmind.com/v1/stable-diffusion-3-medium-txt2img"
 5headers = {
 6    "x-api-key": "YOUR_API_KEY",
 7    "Content-Type": "application/json"
 8}
 9
10data = {
11    "prompt": "A whimsical and high-resolution highly realistic image of a panda in a vintage cosmonaut suit. The panda is holding a sign that reads 'I love flying to the moon!' in playful lettering. The panda's helmet has a small propeller on top and a Indian flag patch, adding to the cosmic vibe. The background features a retro-styled spaceship with rockets and stars, giving the impression of a thrilling journey through space",
12    "negative_prompt": "bad quality, poor quality, doll, disfigured, jpg, toy, bad anatomy, missing limbs, missing fingers, 3d, cgi",
13    "samples": 1,
14    "scheduler": "DPM++ 2M",
15    "num_inference_steps": 25,
16    "guidance_scale": 5,
17    "denoise": 1,
18    "seed": 468685,
19    "img_width": 1024,
20    "img_height": 1024,
21    "modelsamplingsd3_shift": 3,
22    "conditioningsettimesteprange_start": 0.1,
23    "conditioningsettimesteprange_stop": 1,
24    "base64": false
25}
26
27response = requests.post(url, headers=headers, json=data)
28
29if response.status_code == 200:
30    result = response.json()
31    print(json.dumps(result, indent=2))
32else:
33    print(f"Error: {response.status_code}")
34    print(response.text)

API Endpoint

POSThttps://api.segmind.com/v1/stable-diffusion-3-medium-txt2img

Parameters

promptrequired

string

Prompt to render

base64optional

boolean

Base64 encoding of the output image.

Default: false

conditioningsettimesteprange_startoptional

number

Conditioning set timestep range start

Default: 0.1Range: 0.1 - 1

conditioningsettimesteprange_stopoptional

number

Conditioning set timestep range stop

Default: 1Range: 0.1 - 1

denoiseoptional

number

How much to transform the reference image

Default: 1Range: 0.1 - 1

guidance_scaleoptional

number

Scale for classifier-free guidance

Default: 5Range: 1 - 25

img_heightoptional

integer

Image height can be between 512 and 2048 in multiples of 8

Default: 1024

img_widthoptional

integer

Image width can be between 512 and 2048 in multiples of 8

Default: 1024

modelsamplingsd3_shiftoptional

integer

Model Sampling SD3 Shift

Default: 3Range: 1 - 10

negative_promptoptional

string

Prompts to exclude, eg. 'bad anatomy, bad hands, missing fingers'

num_inference_stepsoptional

integer

Number of denoising steps.

Default: 25Range: 10 - 100

samplesoptional

integer

Number of samples to generate.

Default: 1Range: 1 - 4

scheduleroptional

string

Type of scheduler.

Default: "DPM++ 2M"

Allowed values (29 total):

"DPM++ SDE Karras""DPM++ 2M Karras""DPM++ 2M SDE Karras""DPM++ 2M SDE Heun Karras""DPM++ 3M SDE Karras""LMS Karras""DPM2 Karras""DPM2 a Karras""DPM++ 2S a Karras""DPM++ 2M SDE Exponential"+19 more

seedoptional

integer

Seed for image generation.

Default: -1Range: -1 - 999999999999999

Response Type

Returns: Text/JSON

Common Error Codes

The API returns standard HTTP status codes. Detailed error messages are provided in the response body.

400

Bad Request

Invalid parameters or request format

401

Unauthorized

Missing or invalid API key

403

Forbidden

Insufficient permissions

404

Not Found

Model or endpoint not found

406

Insufficient Credits

Not enough credits to process request

429

Rate Limited

Too many requests

500

Server Error

Internal server error

502

Bad Gateway

Service temporarily unavailable

504

Timeout

Request timed out