Installation Guide

This guide covers the installation and initial setup of Prompture.

Requirements

Python 3.8 or higher
pip (Python package installer)

Installation

Installing from PyPI

The recommended way to install Prompture is from PyPI using pip:

pip install prompture

Installing from Source

To install the latest development version from GitHub:

# Clone the repository
git clone https://github.com/your-org/prompture.git
cd prompture

# Install in development mode
pip install -e .

# Or install directly from GitHub
pip install git+https://github.com/your-org/prompture.git

Virtual Environment Setup

It’s recommended to use a virtual environment to avoid dependency conflicts:

# Create a virtual environment
python -m venv prompture-env

# Activate it (Linux/macOS)
source prompture-env/bin/activate

# Activate it (Windows)
prompture-env\Scripts\activate

# Install prompture
pip install prompture

Environment Configuration

Prompture uses environment variables for API configuration. Create a .env file in your project directory:

# OpenAI Configuration
OPENAI_API_KEY=your_openai_api_key_here
OPENAI_BASE_URL=https://api.openai.com/v1  # Optional, uses default if not set

# Anthropic Configuration
ANTHROPIC_API_KEY=your_anthropic_api_key_here

# Google AI Configuration
GOOGLE_API_KEY=your_google_api_key_here

# Groq Configuration
GROQ_API_KEY=your_groq_api_key_here

# Azure OpenAI Configuration
AZURE_OPENAI_API_KEY=your_azure_api_key_here
AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/
AZURE_OPENAI_API_VERSION=2023-12-01-preview

# Local/Custom Endpoints
LOCAL_API_BASE_URL=http://localhost:8000  # For local models
OLLAMA_BASE_URL=http://localhost:11434    # For Ollama

Required API Keys

Depending on which LLM providers you plan to use, you’ll need to obtain API keys:

OpenAI: Get your API key from the OpenAI Dashboard
Anthropic (Claude): Get your API key from the Anthropic Console
Google AI (Gemini): Get your API key from Google AI Studio
Groq: Get your API key from the Groq Console
Azure OpenAI: Set up Azure OpenAI service and get keys from the Azure Portal

Verification

Verify your installation by running a simple test:

import prompture

# Check version
print(f"Prompture version: {prompture.__version__}")

# Test basic import
from prompture import extract_and_jsonify, field_from_registry

print("✅ Prompture installed successfully!")

If you have API keys configured, you can test a basic extraction:

from prompture import extract_and_jsonify

# Simple test (requires valid API key in .env)
try:
    result = extract_and_jsonify(
        prompt="Extract info: John is 25 years old",
        fields={"name": "string", "age": "integer"},
        model_name="openai/gpt-3.5-turbo"  # or your preferred model
    )
    print("✅ API connection successful!")
    print(f"Result: {result}")
except Exception as e:
    print(f"⚠️  API test failed: {e}")
    print("Make sure your API keys are properly configured in .env")

Troubleshooting

Common Installation Issues

ImportError: No module named ‘prompture’

Make sure you’ve installed prompture in the correct Python environment.

API Key Errors

Ensure your .env file is in the correct location and contains valid API keys.

Version Conflicts

If you encounter dependency conflicts, try creating a fresh virtual environment.

Network Issues

If installation fails due to network issues, try:

pip install --upgrade pip
pip install prompture --no-cache-dir

Development Installation

For development work, install with additional development dependencies:

# Clone and install in development mode
git clone https://github.com/your-org/prompture.git
cd prompture
pip install -e ".[dev]"

# Run tests to verify installation
pytest tests/

Docker Installation

You can also use Prompture in a Docker container:

FROM python:3.9-slim

WORKDIR /app
COPY requirements.txt .
RUN pip install prompture

COPY . .
CMD ["python", "your_script.py"]

Next Steps

After installation, check out the Quick Start Guide guide to learn how to use Prompture’s core features, or browse the Examples for practical use cases.

For detailed API documentation, see the API Reference section.