Documentation

This is an example code with fail safes on sending a post request to Modular AI chat completion endpoint:
If you want, you can remove the provider option

import requests

# Replace with your actual API endpoint URL

API_ENDPOINT = 'https://258f3w-8080.csb.app/v1/chat/completions'

# Replace with your actual API key

API_KEY = 'mdl-XXXXXXXXXXXXX'

# Sample data to send in the POST request

data = {

    'model': 'gpt-3.5-turbo', # Choose any model from modular ai model list

    'messages': [

        {"role": "user", "content": "Hello"}

    ],

    'provider': 'A valid Provider name'  # Replace with the provider you wish to use, or remove this line for the default provider

}

# Headers and fail safe

headers = {

    'API-Key': API_KEY,

    'Content-Type': 'application/json'

}

try:

    # Sending a POST request to the API endpoint

    response = requests.post(API_ENDPOINT, json=data, headers=headers)

    # Check if the request was successful (status code 200)

    if response.status_code == 200:

        print("Request successful")

        print("Response:")

        print(response.json())

    else:

        print(f"Request failed with status code {response.status_code}")

        print(response.text)

except requests.RequestException as e:

    print(f"Request failed: {str(e)}")

Generating an Image using python for prodia

import requests

# URL of your endpoint

url = 'https://258f3w-8080.csb.app/v1/imagine/prodia'  # Replace with your actual endpoint URL

# Your API key

api_key = 'mdl-XXXXXXXXXXXXXXXX'

# Sample data for the request

data = {

    'prompt': 'ferrari', #add your custom prompt

    'model': 'absolutereality_v181.safetensors [3d9d4d2b]', #choose your model from the model list

    'negative_prompt': 'ugly', #add negative prompt

    'sampler': 'Euler a',

}

# Sending a POST request with the API key in the headers

headers = {

    'API-Key': api_key

}

response = requests.post(url, json=data, headers=headers)

# Displaying the response content

print(response.text)

Generating an image for crayionv3

import requests

# URL of the Flask API endpoint

api_url = "https://modularai.omiiidev.repl.co/v1/imagine/craiyon"  # Replace this with your actual API URL

# JSON payload with model, prompt, and negative_prompt

payload = {

    "model": "photo",

    "negative_prompt": "Ugly, disformed, broken",

    "prompt": "A cool sports car next to a house"

}

# Replace 'YOUR_API_KEY' with your actual API key

headers = {

    "Api-Key": "mdl-XXXXXXXXXXXXXXXXXXXXXX",

    "Content-Type": "application/json"

}

# Send a POST request to the API endpoint with headers

response = requests.get(api_url, json=payload, headers=headers)

# Process the response

if response.status_code == 200:

    response_data = response.json()

    print("Response:")

    print(response_data)

else:

    print(f"Failed to fetch images. Status code: {response.status_code}")

Using imagine v2:

import requests

# endpoint URL

url = "https://d39a247b-cc2f-462d-87e2-1b6edbe8bb2a-00-6bgmqmbcs7o0.janeway.replit.dev/generate_image"  # Replace with your Flask app's URL

# User-provided parameters

user_data = {

    "models": ["ICBINP - I Can't Believe It's Not Photography"], # choose a model

    "nsfw": False,

    "karras": True,

    "seed_variation": 1000,

    "prompt": "super cool app",

    "steps": 30,

    "sampler_name": "k_euler",

    "width": 512,

    "height": 512,

    "cfg_scale": 7

}

# Sending a POST request to the Flask app

response = requests.post(url, json=user_data)

# Checking the response

if response.status_code == 200:

    result = response.json()

    image_url = result.get("image_url")

    if image_url:

        print("Generated Image URL:", image_url)

    else:

        print("Error: No image URL found in the response.")

else:

    print(f"Error: Failed to generate image. Status Code: {response.status_code}")

    print("Response JSON:", response.json())

Text AI:

/v1/models (list models)

/v1/chat/completions (chat completion)

Image AI:

/v1/imagine/prodia (generate images)

/v1/imagine/prodia/models (image generation ai models for prodia)

/v1/imagine/crayion (generate images)

/v1/tss (generate Text to speech)

Image AI:

/v2/imagine/models  (image generation ai models)

/v2/imagine (generate images)

/v2/predict  (generate image)

/v2/predict/models  (get image models)

/v2/dream (generate image)

Use the TTS:

import requests

# Replace 'your_api_key' with the actual API key

api_key = 'your_api_key'

# Base URL of the Flask app

base_url = 'https://258f3w-8080.csb.app/'

# Example: TTS data

tts_data = {

    'text': 'your prompt here'

}

response = requests.post(f'{base_url}/v1/tts', json=tts_data, headers={'API-Key': api_key})

# Check if the request was successful (status code 200)

if response.status_code == 200:

    try:

        # Try to parse the response as JSON

        result = response.json()

        print(result)

    except requests.exceptions.JSONDecodeError:

        # If parsing fails, print the raw response text

        print(response.text)

else:

    # Print an error message if the request was not successful

    print(f'Error: {response.status_code}, {response.text}')

Generate an Image using dalle 3 or studio ghibli 

import requests

url = "https://258f3w-8080.csb.app/v2/dalle/" 

payload = {

    "model": "Your Model Here",

    "prompt": "You Prompt Here"

}

headers = {

    "Content-Type": "application/json",

    "API-Key": "mdl-XXXXXXXXXXXXXXXXXX" # Replace your api key

}

try:

    response = requests.post(url, json=payload, headers=headers)

    print(response.text)

except Exception as e:

    print(f"Error: {e}")

Generate an Image (API Key not added yet)

import requests

base_url = "https://258f3w-8080.csb.app/v2/imagine/" 

# User-provided parameters

user_data = {

    "models": ["ICBINP - I Can't Believe It's Not Photography"],

    "nsfw": False,

    "karras": True,

    "seed_variation": 1000,

    "prompt": "cyborg",

    "steps": 30,

    "sampler_name": "k_euler",

    "width": 640,

    "height": 512,

    "cfg_scale": 7

}

# Sending a POST request

response = requests.post(base_url, json=user_data)

# Checking the response

if response.status_code == 200:

    result = response.json()

    image_url = result.get("image_url")

    if image_url:

        print("Generated Image URL:", image_url)

    else:

        print("Error: No image URL found in the response.")

else:

    print(f"Error: Failed to generate image. Status Code: {response.status_code}")

    print("Response JSON:", response.json())

/v1/chat/completions
API endpoint to message the models

/v1/models

API endpoint to view all models (not all are available)

Endpoint URL:   https://modualrapi.onrender.com/

Command to use Modular API with terminal copilot:

interpreter --api_base https://258f3w-8080.csb.app/ --api_key "your-api-key" --model openai/gpt-3.5-turbo --context_window 300000