Installing Instructor¶

Instructor is a Python library that works with various LLM providers to extract structured outputs. This guide covers installation and setting up API keys for different providers.

Basic Installation¶

Install the core Instructor package with pip:

pip install instructor

Instructor requires Pydantic for defining data models:

pip install pydantic

Setting Up with Different LLM Providers¶

OpenAI¶

OpenAI is the default provider and works out of the box:

pip install instructor

Set up your OpenAI API key:

export OPENAI_API_KEY=your_openai_key

Anthropic (Claude)¶

To use with Anthropic's Claude models:

pip install "instructor[anthropic]"

Set up your Anthropic API key:

export ANTHROPIC_API_KEY=your_anthropic_key

Google Gemini¶

To use with Google's Gemini models:

pip install "instructor[google-generativeai]"

Set up your Google API key:

export GOOGLE_API_KEY=your_google_key

Cohere¶

To use with Cohere's models:

pip install "instructor[cohere]"

Set up your Cohere API key:

export COHERE_API_KEY=your_cohere_key

Mistral¶

To use with Mistral AI's models:

pip install "instructor[mistralai]"

Set up your Mistral API key:

export MISTRAL_API_KEY=your_mistral_key

LiteLLM (Multiple Providers)¶

To use LiteLLM for accessing multiple providers:

pip install "instructor[litellm]"

Set up API keys for the providers you want to use.

Verifying Your Installation¶

You can verify your installation by running a simple extraction:

import instructor
from openai import OpenAI
from pydantic import BaseModel

class Person(BaseModel):
    name: str
    age: int

client = instructor.from_openai(OpenAI())
person = client.chat.completions.create(
    model="gpt-3.5-turbo",
    response_model=Person,
    messages=[
        {"role": "user", "content": "John Doe is 30 years old"}
    ]
)

print(f"Name: {person.name}, Age: {person.age}")

Next Steps¶

Now that you've installed Instructor, you can:

1. Create your first extraction with a simple model
2. Understand the different response models available
3. Set up clients for your preferred LLM provider

Check the Client Setup guide to learn how to configure clients for different providers.