This is my work in the past two months, bringing new capabilities alongside numerous bug fixes, stability improvements, and quality-of-life enhancements.
✨ Highlights
🌐 Web Search Mode
- Introduce web-powered retrieval with flexible operation modes:
- Combined Web Search + local RAG retrieval
- Web Search only mode for internet-focused research
📑 Text Grounding in Documents
Generated responses can now be grounded directly to document passages, making it easier to verify information and understand where answers come from.
⚡ vLLM Provider Support
Added support for vLLM as an LLM provider.
🛠️ Dev Container Support
Support for development containers, making onboarding and environment consistency significantly easier across teams.
📝 Related project
You may also be interested in HarinezumIgel-LLM-Stack, > a companion project for orchestrating LiteLLM and vLLM.
📦 Guided Installation
A new setup.py installation wizard helps users get started quickly:
- Windows and Unix platforms
- Docker deployments
- Native host installations
🔧 Additional Improvements
This release also includes numerous bug fixes, smaller enhancements, and overall stability improvements across the platform.
