Python SDK

Website | Docs | Blog | Discord

The VLM Run Python SDK is the official Python client for VLM Run API platform, providing a convenient way to interacting with our REST APIs.

🚀 Getting Started

Installation

pip install vlmrun

Installation with Optional Features

The package provides optional features that can be installed based on your needs:

• Video processing features (numpy, opencv-python):

  pip install "vlmrun[video]"

• Document processing features (pypdfium2):

  pip install "vlmrun[doc]"

• All optional features:

  pip install "vlmrun[all]"

Basic Usage

from vlmrun.client import Client
from vlmrun.hub.schemas.document.invoice import Invoice

# Initialize the client
client = Client(api_key="your-api-key")

# Process an image
response = client.image.generate(
    image="https://example.com/invoice.jpg",
    model="vlm-1",
    domain="document.invoice",
    json_schema=Invoice.model_json_schema(),
)

Image Utilities

from vlmrun.common.image import encode_image
from vlmrun.common.utils import download_image
from PIL import Image

# Convert image to base64 or binary
image = Image.open("image.jpg")
base64_str = encode_image(image, format="PNG")  # or format="JPEG"
binary_data = encode_image(image, format="binary")

# Download image from URL
image = download_image("https://example.com/image.jpg")

📂 Directory Structure

vlmrun/
├── client/               # Client implementation
│   ├── client.py         # Main Client class
│   ├── base_requestor.py # Low-level request logic
│   ├── files.py          # File operations
│   ├── models.py         # Model operations
│   ├── finetune.py       # Fine-tuning operations
│   └── types.py          # Client type definitions
├── common/              # Common utilities
│   ├── auth.py          # Authentication utilities
│   └── image.py         # Image processing utilities
└── types/              # Type definitions
    └── abstract.py     # Abstract base classes

🔗 Quick Links

• 💬 Need help? Email us at [email protected] or join our Discord
• 📚 Check out our Documentation
• 📣 Follow us on Twitter and LinkedIn