Python – How to use the Stablediffusion API in 4 easy steps

Step 1 – Generate a bearer token

Login and then create an application from the My account menu.

Generate a bearer token from your new application.

Step 2 – Start a new thread with Stablediffusion

You may select an ai_version_code and an engine. For example : stablediffusion_text2img and stable-diffusion-xl-1024-v0-9.

Send an http POST request to create a new thread.

import requests
import time

bearer = '<<your_token_here>>'

# CREATE THREAD
# -------------
url = "https://app.ai-client.com/api/v1/threads"

payload={'ai_version_code': 'stablediffusion_text2img'}

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("POST", url, headers=headers, data=payload)

thread = response.json()['data']

print(thread)

This will return :

{
   "guid":"44142638-4934-4bf5-9be7-b4b5a6368f44",
   "ai_code":"stablediffusion",
   "ai_version_code":"stablediffusion_text2img",
   "title":"New thread with Stable Diffusion Image Generation",
   "created_at":"2023-07-13T19:56:25.000000Z",
   "updated_at":"2023-07-13T19:56:25.000000Z"
}

Step 3 – Create a thread entry to send a message

You may use the thread object to create a new entry with a POST request.

url = "https://app.ai-client.com/api/v1/threads/" + thread['guid'] + "/entry"

payload={
  'type': 'request',
  'engine' : 'stable-diffusion-xl-1024-v0-9',
  'resolution' : '1024x1024',
  # positive prompt: what we want to see
  'prompts[0][text]' : "Delorean hacking time",
  'prompts[0][weight]' : 1,
  # negative prompt: what we want to avoid
  'prompts[1][text]' : "Clock",
  'prompts[1][weight]' : -1,
  'images' : 2,
  'steps' : 45,
  'style' : "neon-punk"
}

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("POST", url, headers=headers, data=payload)

thread_entry = response.json()['data']

print(thread_entry)

This will return :

{
   "guid":"c7c088d3-7644-4385-b77c-a3ee98c4eed2",
   "thread_guid":"44142638-4934-4bf5-9be7-b4b5a6368f44",
   "type":"request",
   "error":"None",
   "content":{
      "prompts":[
         {
            "text":"Delorean hacking time",
            "weight":1
         },
         {
            "text":"Clock",
            "weight":-1
         }
      ],
      "engine":"stable-diffusion-xl-1024-v0-9",
      "style":"neon-punk",
      "images":2,
      "steps":45,
      "resolution":"1024x1024"
   },
   "waiting_response":true,
   "credits":0,
   "created_at":"2023-07-13T19:56:25.000000Z"
}

Step 4 – Wait for a response from Stablediffusion

You must now wait for the AI’s response. The “waiting_response” property of your thread entry informs you of Stablediffusion’s response.

You may use the thread object to get all the thread entries. Send a GET request. You will see your thread entry of type “request” and the Stablediffusion thread entry of type “response“.

url = "https://app.ai-client.com/api/v1/threads/" + thread['guid'] + "/entries"

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("GET", url, headers=headers)

thread_entries = response.json()['data']

This will return :

[
   {
      "guid":"c7c088d3-7644-4385-b77c-a3ee98c4eed2",
      "thread_guid":"44142638-4934-4bf5-9be7-b4b5a6368f44",
      "type":"request",
      "error":"None",
      "content":{
         "prompts":[
            {
               "text":"Delorean hacking time",
               "weight":1
            },
            {
               "text":"Clock",
               "weight":-1
            }
         ],
         "engine":"stable-diffusion-xl-1024-v0-9",
         "style":"neon-punk",
         "images":2,
         "steps":45,
         "resolution":"1024x1024"
      },
      "waiting_response":false,
      "credits":0,
      "created_at":"2023-07-13T19:56:25.000000Z"
   },
   {
      "guid":"a9e17dcc-a162-41ed-a1b4-94ecc40cdf90",
      "thread_guid":"44142638-4934-4bf5-9be7-b4b5a6368f44",
      "type":"response",
      "error":"None",
      "content":[
         {
            "url":"https://app.ai-client.com/storage/stablediffusion_text2img/2212c664-7d5b-4fe5-b240-1f5e46e2e70e.png",
            "thumbnail_url":"https://app.ai-client.com/storage/stablediffusion_text2img/2212c664-7d5b-4fe5-b240-1f5e46e2e70e.png.thumbnail.jpg",
            "dimensions":"1024x1024",
            "filesize":1409734
         },
         {
            "url":"https://app.ai-client.com/storage/stablediffusion_text2img/e65893b4-acfb-459d-9ee4-796d724b9689.png",
            "thumbnail_url":"https://app.ai-client.com/storage/stablediffusion_text2img/e65893b4-acfb-459d-9ee4-796d724b9689.png.thumbnail.jpg",
            "dimensions":"1024x1024",
            "filesize":1685147
         }
      ],
      "waiting_response":false,
      "credits":10,
      "created_at":"2023-07-13T19:57:09.000000Z"
   }
]

Full code example

import requests
import time

bearer = '<<your_token_here>>'

# CREATE THREAD
# -------------
url = "https://app.ai-client.com/api/v1/threads"

payload={'ai_version_code': 'stablediffusion_text2img'}

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("POST", url, headers=headers, data=payload)

thread = response.json()['data']

print(thread)
# {'guid': '44142638-4934-4bf5-9be7-b4b5a6368f44', 'ai_code': 'stablediffusion', 'ai_version_code': 'stablediffusion_text2img', 'title': 'New thread with Stable Diffusion Image Generation', 'created_at': '2023-07-13T19:56:25.000000Z', 'updated_at': '2023-07-13T19:56:25.000000Z'}


# CREATE THREAD ENTRY OF TYPE request
# -----------------------------------
url = "https://app.ai-client.com/api/v1/threads/" + thread['guid'] + "/entry"

payload={
  'type': 'request',
  'engine' : 'stable-diffusion-xl-1024-v0-9',
  'resolution' : '1024x1024',
  # positive prompt: what we want to see
  'prompts[0][text]' : "Delorean hacking time",
  'prompts[0][weight]' : 1,
  # negative prompt: what we want to avoid
  'prompts[1][text]' : "Clock",
  'prompts[1][weight]' : -1,
  'images' : 2,
  'steps' : 45,
  'style' : "neon-punk"
}

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("POST", url, headers=headers, data=payload)

thread_entry = response.json()['data']

print(thread_entry)
# {'guid': 'c7c088d3-7644-4385-b77c-a3ee98c4eed2', 'thread_guid': '44142638-4934-4bf5-9be7-b4b5a6368f44', 'type': 'request', 'error': None, 'content': {'prompts': [{'text': 'Delorean hacking time', 'weight': 1}, {'text': 'Clock', 'weight': -1}], 'engine': 'stable-diffusion-xl-1024-v0-9', 'style': 'neon-punk', 'images': 2, 'steps': 45, 'resolution': '1024x1024'}, 'waiting_response': True, 'credits': 0, 'created_at': '2023-07-13T19:56:25.000000Z'}


# WAIT FOR 2 MINUTES
# ------------------
time.sleep(120)

# GET THREAD ENTRIES
# ------------------
url = "https://app.ai-client.com/api/v1/threads/" + thread['guid'] + "/entries"

headers = {
  'Accept': 'application/json',
  'Authorization': 'Bearer ' + bearer
}

response = requests.request("GET", url, headers=headers)

thread_entries = response.json()['data']

print(thread_entries)
# [{'guid': 'c7c088d3-7644-4385-b77c-a3ee98c4eed2', 'thread_guid': '44142638-4934-4bf5-9be7-b4b5a6368f44', 'type': 'request', 'error': None, 'content': {'prompts': [{'text': 'Delorean hacking time', 'weight': 1}, {'text': 'Clock', 'weight': -1}], 'engine': 'stable-diffusion-xl-1024-v0-9', 'style': 'neon-punk', 'images': 2, 'steps': 45, 'resolution': '1024x1024'}, 'waiting_response': False, 'credits': 0, 'created_at': '2023-07-13T19:56:25.000000Z'}, {'guid': 'a9e17dcc-a162-41ed-a1b4-94ecc40cdf90', 'thread_guid': '44142638-4934-4bf5-9be7-b4b5a6368f44', 'type': 'response', 'error': None, 'content': [{'url': 'https://app.ai-client.com/storage/stablediffusion_text2img/2212c664-7d5b-4fe5-b240-1f5e46e2e70e.png', 'thumbnail_url': 'https://app.ai-client.com/storage/stablediffusion_text2img/2212c664-7d5b-4fe5-b240-1f5e46e2e70e.png.thumbnail.jpg', 'dimensions': '1024x1024', 'filesize': 1409734}, {'url': 'https://app.ai-client.com/storage/stablediffusion_text2img/e65893b4-acfb-459d-9ee4-796d724b9689.png', 'thumbnail_url': 'https://app.ai-client.com/storage/stablediffusion_text2img/e65893b4-acfb-459d-9ee4-796d724b9689.png.thumbnail.jpg', 'dimensions': '1024x1024', 'filesize': 1685147}], 'waiting_response': False, 'credits': 10, 'created_at': '2023-07-13T19:57:09.000000Z'}]

See our Swagger API documentation for more information.