Strengths
- Supports self-hosting via Docker or direct install, keeping data local.
- Handles multimodal documents natively, embedding full pages for accurate search without parsing loss.
- Free for personal and commercial use under $2,000/month revenue with easy model configuration.
- Includes console UI, API documentation, and Model Context Protocol for querying without coding.
- Stores data persistently in Docker volumes and local directories for documents and logs.
Limitations
- No full support for self-hosted deployments; relies on guides and Discord community for assistance.
- Initial Docker setup downloads large AI models requiring 10GB disk space and good internet connection, taking 5-10 minutes.
- Commercial use over $2,000/month revenue requires a paid key; free cloud plan quotas are limited.