LibreChat Search Stack Generator

🎥 Demo (Coming Soon)

▶️ Try the live demo - Generate your local search stack in seconds!

🔗 Side Tool for LibreChatConfigurator

This is a companion tool for LibreChatConfigurator that simplifies setting up a local search stack for LibreChat.

Main Tool: LibreChatConfigurator - Configure all LibreChat settings (AI providers, OAuth, database, etc.)

This Tool: Local Search Stack Generator - Generate Docker containers for LibreChat's search capabilities when you need them

How They Work Together:

⚙️ Start with LibreChatConfigurator - Configure your LibreChat instance
- Set up AI providers (OpenAI, Anthropic, etc.)
- Configure authentication and OAuth
- Set up database and file storage
🔍 When you need local search - Use this tool to generate the Docker stack
- Configure SearXNG, Jina Reader, and BGE Reranker services
- Download the generated package with Docker Compose files
- Run the installation script to deploy services
🔄 Import configuration back - Auto-configure LibreChat's search settings
- In LibreChatConfigurator: Configuration → Import Merge JSON
- Or directly under Search section, use the import function
- Your local search stack is now connected to LibreChat!

Complete Workflow:

LibreChatConfigurator → Need Local Search? → Generate Stack → Run Install → Import JSON Config
   (main tool)              (this tool)         (Docker)      (back to main tool)

Why This Exists

LibreChat's local search stack requires three Docker services working in harmony, but setting them up shouldn't require hours of Docker debugging.

Configuring LibreChat's web search feature (available in v0.7.9+) involves deploying SearXNG for meta-search, Jina Reader for web scraping, and BGE Reranker for ML-powered result ranking. This requires:

Finding working Docker images (many community images are outdated or broken)
Configuring complex networking between containers and LibreChat
Setting up environment variables, ports, and resource limits correctly
Writing Docker Compose files with health checks and restart policies
Creating cross-platform installation scripts
Manually configuring LibreChat's YAML to connect to these services

This tool solves that. Whether you're setting up LibreChat for your own use or deploying for others, you get:

Working Docker Images - Pre-tested, community-maintained images (no broken containers)
Docker Networking - Automatic shared network setup for LibreChat container communication
1-Click Installation - Complete package with Docker Compose, scripts for Windows/Mac/Linux
Auto-Configuration - JSON config that imports directly into LibreChatConfigurator
Resource Optimized - Pre-configured memory limits and health checks for stable operation
Beginner-Friendly - No Docker or YAML knowledge required

Built by the community, for the community. As LibreChat adds new search features, this tool evolves with it. Missing a configuration option or want to add a new service? Contributions welcome!

Help make LibreChat's search accessible to everyone. Whether you're fixing a bug, improving Docker configurations, or enhancing the user experience, your contributions help more people deploy powerful local search without the setup headaches.

Quick Start - Use Online

When you reach the search configuration step in LibreChatConfigurator, follow this workflow:

🚀 1-Minute Setup

Open the online generator: https://librechatlocalwebsearchstack.netlify.app/
Configure your services (optional - defaults work great!)
- Adjust ports if needed
- Set resource limits for your system
- Click "Generate & Download"

Install the Docker stack

# Extract the downloaded ZIP
unzip librechat-search-stack.zip
cd librechat-search-stack

# Run the installation script
./install_dockerimage.sh    # Unix/Linux/Mac
# OR
install_dockerimage.bat     # Windows

Import into LibreChatConfigurator
- Open the extracted folder and find search-stack-config.json
- In LibreChatConfigurator: Configuration → Import Merge JSON
- Or use the import function directly under the Search section
- Select the search-stack-config.json file
- Done! Your search settings are auto-configured 🎉

✨ What You Get

✅ No installation required - runs entirely in your browser
✅ Configure all search service settings with live preview
✅ Download complete ZIP package with all configuration files
✅ Export JSON configuration for LibreChatConfigurator import
✅ 1-click installation scripts for all platforms

💻 Run Locally (Alternative)

Prerequisites: Node.js 20+ is required

📦 Install Node.js (click to expand)

Windows:

# Download and run installer from nodejs.org
# Or use winget:
winget install OpenJS.NodeJS

macOS:

# Using Homebrew (recommended):
brew install node

# Or download installer from nodejs.org

Linux (Ubuntu/Debian):

# Install Node.js 20:
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs

Verify installation (all platforms):

node --version   # Should show v20.x.x or higher
npm --version    # Should show npm version

Run Locally

# Clone the repository
git clone https://github.com/YOUR_USERNAME/librechat-search-stack.git
cd librechat-search-stack

# Install dependencies
npm install

# Start the development server
npm run dev

# Open http://localhost:5000 in your browser

Same features as the online version, but running on your own machine.

Production Build

For production deployment:

npm run build
npm start

The app will build and serve on port 5000 in production mode.

Features

3 Essential Search Services: Configure SearXNG, Jina Reader (Firecrawl alternative), and BGE Reranker v2-m3
Tested Docker Images: Uses working community-maintained images (no broken containers)
Live Preview: See docker-compose.yml, .env, README.md, and install scripts in real-time
JSON Export: Generate search-stack-config.json that imports directly into LibreChatConfigurator
Complete Package: Download ZIP with all 5 files ready for deployment
1-Click Installation: Bash script included for automatic Docker deployment

What's Included in the Package

When you click "Generate & Download", you get a ZIP file containing:

📄 Generated Files

File	Description
`docker-compose.yml`	Complete Docker Compose configuration with all 3 services
`.env`	Pre-configured environment variables (ready to use, no editing needed)
`README.md`	Setup instructions and service documentation
`install_dockerimage.sh`	Unix/Linux/Mac installation script
`install_dockerimage.bat`	Windows installation script
`search-stack-config.json`	LibreChat search configuration - Import into LibreChatConfigurator
`test_services.py`	Python test script to verify all services are working

🐳 Docker Services

Service	Docker Image	Purpose	Default Port
SearXNG	`searxng/searxng:latest`	Meta-search engine (Google, Bing, etc.)	8080
Jina Reader	`ghcr.io/intergalacticalvariable/reader`	Web scraping & content extraction	3000
BGE Reranker	`wkao/bge-reranker-v2-m3`	ML-based result ranking	8787

LibreChat Compatibility

Supported Versions

This tool generates configurations compatible with LibreChat v0.7.9+ (web search feature introduced in v0.7.9).

Service Components

LibreChat's web search requires three components working together:

Component	Role	This Tool Provides	LibreChat Config Mapping
Search Provider	Performs web searches	SearXNG (`latest`)	`searchProvider: "searxng"`
Scraper	Extracts content from URLs	Jina Reader (Firecrawl-compatible)	`scraperType: "firecrawl"`
Reranker	ML-powered result ranking	BGE Reranker v2-m3 (Jina API-compatible)	`rerankerType: "jina"`

Technical Details

SearXNG (Search Engine)

Version: Latest stable release
API Format: JSON output (format=json)
Features: Privacy-focused meta-search aggregating 70+ search engines
Network: Supports X-Forwarded-For headers for query logging
Authentication: Configurable API key (default: searxng-default-key-12345)

Jina Reader (Web Scraper / Firecrawl Alternative)

Version: Latest from ghcr.io/intergalacticalvariable/reader
Compatibility: Drop-in Firecrawl replacement
API: RESTful endpoint at http://service:3000/{url_to_crawl}
Features: Clean content extraction, JavaScript rendering
Configuration: Timeout (30s default), max pages limit
Authentication: Configurable API key (default: jina-default-key-67890)

BGE Reranker v2-m3 (Result Ranking)

Version: BAAI/bge-reranker-v2-m3 (278M parameters)
Docker Image: wkao/bge-reranker-v2-m3 (community-maintained)
API Compatibility: Jina Reranker API-compatible endpoint
Endpoint: POST /api/v1/rerank with JSON payload
Features: Multilingual support (100+ languages), optimized for RAG
Performance: Batch processing up to 16 documents
Authentication: Configurable API key (default: reranker-default-key-abcde)

Docker Network Architecture

Shared Network: All services join external Docker network (default: librechat)
Container DNS: Services accessible via container names (searxng:8080, jina-reader:3000, bge-reranker:8787)
LibreChat Integration: Import search-stack-config.json to auto-configure container URLs
Host Access: Services also available via localhost URLs for testing

📚 LibreChat Web Search Documentation

Integration Workflow

The Recommended Way to Set Up LibreChat with Local Search

Start with LibreChatConfigurator (the main configuration tool), then use this tool when you need local search capabilities.

Step-by-Step

🔧 Configure LibreChat (LibreChatConfigurator)
- Run LibreChatConfigurator
- Set up AI providers (OpenAI, Anthropic, etc.)
- Configure authentication and OAuth
- Set up database and file storage
🔍 Need Local Search? Use this tool to generate the Docker stack
- Open https://librechatlocalwebsearchstack.netlify.app/
- Configure service ports and resource limits (or use defaults)
- Click "Generate & Download" to get the ZIP package
- Extract and run the installation script (./install_dockerimage.sh or .bat)
🔄 Import Search Config Back to LibreChatConfigurator
- In LibreChatConfigurator: Configuration → Import Merge JSON
- Or use the import function directly under the Search section
- Select search-stack-config.json from the downloaded ZIP
- Search settings auto-populate with your Docker service URLs
🚀 Deploy Everything
- Docker search stack runs on your configured ports
- LibreChat connects to local search services via the imported config
- Complete AI-powered search infrastructure ready!

Configuration Mapping

The search-stack-config.json uses LibreChat's official configuration structure:

{
  "name": "Search Stack Configuration",
  "configuration": {
    "webSearch": {
      "searchProvider": "searxng",
      "searxngInstanceUrl": "http://localhost:8080",
      "scraperType": "firecrawl",
      "firecrawlApiUrl": "http://localhost:3000",
      "rerankerType": "jina",
      "jinaRerankerUrl": "http://localhost:8787",
      "safeSearch": true
    }
  }
}

Note: Ports match your Docker configuration. LibreChatConfigurator reads these values and configures LibreChat to connect to your local search stack.

For detailed service specifications, see LibreChat Compatibility above.

Architecture

This project follows the same architecture as LibreChatConfigurator for consistency:

LibreChat Search Stack Generator
├── client/                 # React Frontend
│   ├── src/
│   │   ├── components/     # Service Cards, Preview Panel
│   │   ├── lib/           # File Generation Logic
│   │   ├── pages/         # Home Page
│   │   └── hooks/         # Configuration State
├── server/                 # Express Backend (minimal)
│   ├── routes.ts          # No API routes needed (client-side generation)
│   └── vite.ts            # Development Server
├── shared/                 # Shared Types & Schemas
│   └── schema.ts          # Service Configuration Schemas
└── README.md              # This file

Data Flow

User Input → Service configuration forms (ports, resources)
Client-Side Generation → JSZip creates files in browser
File Download → User gets complete package with all files
Docker Deployment → User runs installation script
LibreChat Integration → Import JSON into LibreChatConfigurator

Key Technical Decisions

Client-Side File Generation: All files generated in browser using JSZip (no backend needed)
No Database Required: Simple configuration tool, no data persistence
Reusable Components: Service cards and preview panels
Type Safety: Zod schemas ensure valid configurations

For Developers

Frontend: /client - React with TypeScript, Tailwind CSS, and shadcn/ui components
File Generation: /client/src/lib/file-generator.ts - Docker Compose, env, JSON, and script generation
Shared: /shared - Service configuration schemas
Presets: /client/src/lib/presets.ts - Default service configurations

Development Commands

npm run dev          # Start development server
npm run build        # Production build
npm run start        # Start production server

Adding New Services

Update Schema: Add new service to /shared/schema.ts
Add Preset: Configure defaults in /client/src/lib/presets.ts
Update Generator: Modify /client/src/lib/file-generator.ts to include new service
Add UI Card: Create service configuration card in /client/src/components/
Update JSON Export: Include new service settings in JSON configuration

Contributing

We welcome contributions! This tool complements LibreChatConfigurator and should evolve together.

Quick Contribution Steps

Fork the repository
Create a feature branch: git checkout -b feature/new-service
Make your changes and test thoroughly
Submit a pull request with a clear description

Areas We Need Help

New Search Services: Add support for additional search engines or scrapers
Docker Image Updates: Keep track of working community images
UI/UX Improvements: Enhance configuration experience
Documentation: Improve setup guides and integration docs
Testing: Add comprehensive test coverage

Related Projects

LibreChatConfigurator - Complete LibreChat configuration tool
LibreChat - Enhanced ChatGPT clone with multiple AI providers

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

LibreChat Team - For creating an amazing open-source AI chat platform
Community Docker Maintainers - For keeping Jina Reader and BGE Reranker images working
SearXNG Project - For the privacy-focused meta-search engine
LibreChatConfigurator - For inspiring the configuration tool approach

Name		Name	Last commit message	Last commit date
Latest commit History 61 Commits
attached_assets		attached_assets
client		client
server		server
shared		shared
.gitignore		.gitignore
.replit		.replit
README.md		README.md
components.json		components.json
drizzle.config.ts		drizzle.config.ts
librechat-search-stack.zip		librechat-search-stack.zip
package-lock.json		package-lock.json
package.json		package.json
postcss.config.js		postcss.config.js
replit.md		replit.md
search-stack-config.json		search-stack-config.json
tailwind.config.ts		tailwind.config.ts
tsconfig.json		tsconfig.json
vite.config.ts		vite.config.ts

Folders and files

Latest commit

History

Repository files navigation

LibreChat Search Stack Generator

🎥 Demo (Coming Soon)

🔗 Side Tool for LibreChatConfigurator

Why This Exists

Quick Start - Use Online

🚀 1-Minute Setup

✨ What You Get

💻 Run Locally (Alternative)

Run Locally

Production Build

Features

What's Included in the Package

📄 Generated Files

🐳 Docker Services

LibreChat Compatibility

Supported Versions

Service Components

Technical Details

Docker Network Architecture

Integration Workflow

The Recommended Way to Set Up LibreChat with Local Search

Step-by-Step

Configuration Mapping

Architecture

Data Flow

Key Technical Decisions

For Developers

Development Commands

Adding New Services

Contributing

Quick Contribution Steps

Areas We Need Help

Related Projects

License

Acknowledgments

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages