Download the PHP package darvis/laravel-ai-generator without Composer

Laravel AI Generator

darvis/laravel-ai-generator writes structured content for a Laravel application with Claude, ChatGPT, Gemini or Grok. You pass a ContentRequest with a topic and get a ContentResult with a title, an intro, the main text as HTML, an SEO title, a meta description and optionally an image.

Features

• One call returns title, intro, text (HTML with h2, p, ul and li), seoTitle and seoDescription.
• Four providers out of the box: OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini) and xAI (Grok). Every one is asked for strict JSON, so you get separate fields instead of one text to parse.
• Use several at once: one provider writes, another makes the images (Claude has no image model), and others take over when the first one fails. Pick another provider or model per call with AiGenerator::using('anthropic').
• A setup wizard, php artisan ai-generator:install, that checks every API key with the provider, lets you pick the models your key may use and writes .env for you. php artisan ai-generator:status checks the keys again at no cost.
• An MCP server (with laravel/mcp), so an assistant such as Claude Code can generate content and images through your application.
• Language, tone, reading level, audience, keywords, brand, call to action and length per request, with defaults in the config.
• A failing image does not lose the text: the result comes back with hasError() true.
• No routes, views or migrations: you call it from your own code and store the result yourself.
• Ships a Laravel Boost guideline and skill, so an AI assistant in your application knows the API.

Requirements

• PHP 8.2 or higher (Laravel 13 itself needs PHP 8.3)
• Laravel 11, 12 or 13
• An API key of at least one provider. A ChatGPT, Claude or Gemini subscription is not an API key: the API is billed separately, per call. The API keys guide shows step by step where to create one.
• Optional: laravel/mcp (Laravel 12.41 or newer) for the MCP server

Installation

The wizard asks which providers you have a key for, checks each key, and writes the settings to .env. You can also set them by hand:

Never commit .env or put a key in your code. Publishing the config file is optional:

Quick start

Another provider or model for one call:

includeImage is true by default. Leave out includeImage: false and the package also generates an image, which is a second, billed call. The text field is HTML written by a model: sanitise it before you render it unescaped.

Documentation

The full documentation lives on the documentation site:

• API keys: where to create a key for each provider, and how to keep it secret
• Installation: the wizard step by step, with a check that costs nothing
• Multiple providers: default, image and fallback drivers, and using()
• MCP server: let an assistant generate content through your app
• Usage: a complete example, every request option, images, errors and queued jobs
• Configuration: every option and environment variable
• Custom drivers: let another AI provider write the text
• Testing: test your code without calling a real API
• Troubleshooting: every error message with cause and fix
• API reference: every public class and method
• FAQ: the short answers

Laravel Boost

The package ships a Laravel Boost guideline and skill. Run php artisan boost:install, or php artisan boost:update --discover in a project that already uses Boost.

Testing

Changelog

See CHANGELOG.

Support the package

If darvis/laravel-ai-generator saves you time, a star on GitHub or a favourite on Packagist helps other developers find it.

Contributing

See CONTRIBUTING.

Security

Please report a vulnerability privately, as described in SECURITY, not in the issue tracker.

License

The MIT License (MIT). See LICENSE.