This plugin enhances Payload CMS with AI-powered content creation and management features, targeting developers and content creators looking to streamline workflows. It offers AI assistance for text generation, proofreading, translation, and voice/image generation, aiming to boost productivity and content quality.

How It Works

The plugin integrates with Payload CMS by adding AI capabilities directly into content fields, particularly Text and RichText fields via Lexical editor features. It supports multiple AI providers (OpenAI, Anthropic, ElevenLabs) through environment variable configuration, allowing users to bring their own models. Key features include field-level prompt customization and access control for AI settings.

Quick Start & Requirements

• Install via pnpm add @ai-stack/payloadcms.
• Requires Payload CMS v3.31.0 or later.
• Environment variables for API keys (OPENAI_API_KEY, ANTHROPIC_API_KEY, ELEVENLABS_API_KEY) are necessary.
• Server restart required after configuration.
• Official Guide: https://github.com/ashbuilds/payload-ai/blob/main/docs/guide.md
• Demo: https://github.com/ashbuilds/payload-ai/blob/main/docs/demo.md

Highlighted Details

• Supports Text and RichText fields for AI operations.
• Integrates with ElevenLabs for voice generation and OpenAI for image generation.
• Offers field-level prompt customization and access control.
• Allows BYOM (Bring Your Own Model) configuration.

Maintenance & Community

• Actively developed with regular updates.
• Community engagement encouraged via Payload's Discord.

Licensing & Compatibility

• License not explicitly stated in the README.
• Compatibility with Payload CMS v3.31.0 is confirmed; broader compatibility is under development.

Limitations & Caveats

The plugin is in active development and has only been tested with Payload v3.31.0, indicating potential compatibility issues with other versions. Features like summarization, expansion, fact-checking, and automated workflows are marked as "Coming Soon" or not yet implemented. Custom field integration for AI features may require manual component path injection.