Skip to main content
This guide will help you get started with Instructor in your PHP project in under 5 minutes. For detailed setup instructions, see Setup.

​
Install Instructor with Composer

Run following command in your terminal: 
composer require cognesy/instructor-php
# @doctest id="00c4"

​
Create and Run Example

​
Step 1: Prepare your OpenAI API Key

In this example, we’ll use OpenAI as the LLM provider. You can get it from the OpenAI dashboard.

​
Step 2: Create a New PHP File

In your project directory, create a new PHP file test-instructor.php: 
<?php
require __DIR__ . '/vendor/autoload.php';

use Cognesy\Instructor\StructuredOutput;

// Set up OpenAI API key
$apiKey = 'your-openai-api-key';
putenv("OPENAI_API_KEY=" . $apiKey);
// WARNING: In real project you should set up API key in .env file.

// Step 1: Define target data structure(s)
class City {
    public string $name;
    public string $country;
    public int $population;
}

// Step 2: Use Instructor to run LLM inference
$city = (new StructuredOutput)
    ->using('openai')
    ->withResponseClass(City::class)
    ->withMessages('What is the capital of France?')
    ->get();

var_dump($city);
// @doctest id="528c"
You should never put your API keys directly in your real project code to avoid getting them compromised. Set them up in your .env file.

​
Step 3: Run the Example

Now, you can run the example: 
php test-instructor.php

# Output:
# object(City)#1 (3) {
#   ["name"]=>
#   string(5) "Paris"
#   ["country"]=>
#   string(6) "France"
#   ["population"]=>
#   int(2148000)
# }
# @doctest id="9cbb"

​
Next Steps

You can start using Instructor in your project right away after installation. But it’s recommended to publish configuration files and prompt templates to your project directory, so you can customize the library’s behavior and use your own prompt templates. You should also set up LLM provider API keys in your .env file instead of putting them directly in your code. See Setup Instructions for more details.
⌘I