TurboStream

February 6, 2026 · View on GitHub

Real-time data streaming backend with AI-powered analysis

High-performance Go backend providing REST APIs and WebSocket streaming for monitoring high-velocity data streams with LLM analysis.

Quick Start Video

Coming Soon: Watch our quick start guide on YouTube

Get Started in 3 Steps

1. Clone & Configure

git clone https://github.com/turboline-ai/turbostream.git
cd turbostream
cp .env.local.example .env.local

2. Set Up Your Environment

Edit .env.local with your credentials:

# Database (Required)
MONGODB_URI=your-mongodb-connection-string
MONGODB_DB_NAME=turbostream

# Security (Required)
JWT_SECRET=$(openssl rand -hex 64)
ENCRYPTION_KEY=$(openssl rand -hex 32)

# LLM Provider (Choose one)
DEFAULT_AI_PROVIDER=anthropic
ANTHROPIC_API_KEY=sk-ant-your-key

Need MongoDB? Get a free cluster at mongodb.com/cloud/atlas

Need an LLM API key? See LLM Providers below

3. Run the Server

go run ./cmd/server

Server starts at http://localhost:7210

LLM Providers

TurboStream supports Bring Your Own Model (BYOM). Configure any provider by setting environment variables:

Provider	Get API Key	Env Vars
Anthropic (Claude)	console.anthropic.com	`ANTHROPIC_API_KEY`, `ANTHROPIC_MODEL`
OpenAI	platform.openai.com	`OPENAI_API_KEY`, `OPENAI_MODEL`
Azure OpenAI	portal.azure.com	`AZURE_OPENAI_*`
Google Gemini	aistudio.google.com	`GOOGLE_API_KEY`, `GOOGLE_MODEL`
Mistral	console.mistral.ai	`MISTRAL_API_KEY`, `MISTRAL_MODEL`
xAI (Grok)	console.x.ai	`XAI_API_KEY`, `XAI_MODEL`
Ollama (Local)	ollama.com	`OLLAMA_BASE_URL`, `OLLAMA_MODEL`

Example (Anthropic):

DEFAULT_AI_PROVIDER=anthropic
ANTHROPIC_API_KEY=sk-ant-your-key-here
ANTHROPIC_MODEL=claude-3-5-sonnet-20241022

Deploy to Production

Railway (Recommended)

Coming Soon: Watch our Railway deployment guide

Push your code to GitHub
Connect to Railway: railway.app
Add environment variables in Railway dashboard
Deploy - Railway auto-detects your Dockerfile

Need help? See our deployment guide

Docker

# Build
docker build -t turbostream .

# Run
docker run -p 7210:7210 --env-file .env.local turbostream

API Documentation

Interactive API Documentation

Try it now: When your server is running, visit:

http://localhost:7210/docs

The API documentation (powered by Scalar) lets you:

Explore all API endpoints interactively
Test authentication, marketplace, LLM, and settings APIs
View request/response schemas with examples
Execute API calls directly from your browser

ℹ️ Note: If you're using an ad-blocker (e.g., Brave Shields, uBlock Origin), you may need to disable it for localhost:7210 to test API calls in the browser.

⚠️ WebSocket Limitation: WebSocket streaming (/ws) cannot be fully tested in the browser documentation. For WebSocket testing, use CLI tools like wscat or Postman with WebSocket support.

Quick Reference

Authentication:

POST /api/auth/register - Register new user
POST /api/auth/login - Login and get JWT token
GET /api/auth/me - Get current user (requires auth)

LLM:

POST /api/llm/query - Query LLM with feed data
POST /api/llm/query/stream - Stream LLM responses (SSE)
GET /api/llm/providers - List available LLM providers

Marketplace:

GET /api/marketplace/feeds - List all feeds
POST /api/marketplace/feeds - Create new feed (requires auth)
POST /api/marketplace/subscribe/:feedId - Subscribe to feed
GET /api/marketplace/subscriptions - Get user subscriptions

WebSocket:

ws://localhost:7210/ws - Real-time streaming

Complete Documentation: turboline.ai/docs/api

Features

Core Capabilities

JWT Authentication - Secure user management with 2FA support
WebSocket Streaming - Real-time data and LLM token streaming
Multi-Provider LLM - Support for 7+ LLM providers (BYOM)
Feed Marketplace - Discover and manage data feeds
Topic Routing - Isolated LLM intelligence per topic (e.g., BTC-USD, ETH-USD)
Context Management - Intelligent data accumulation for AI analysis
Token Optimization - TSLN format reduces tokens by 40-60%

Security

JWT-based authentication
2FA with TOTP and backup codes
Encrypted sensitive data
Session management
Login activity tracking

Clients

Terminal UI

A keyboard-driven terminal interface for TurboStream:

Repository: github.com/turboline-ai/turbostream-tui

git clone https://github.com/turboline-ai/turbostream-tui.git
cd turbostream-tui
go run .

Web Frontend (Commercial)

Modern web interface with responsive design and user management.

Contact: turboline.ai for commercial licensing

Development

Project Structure

turbostream/
├── cmd/server/          # Main application entry point
├── internal/
│   ├── config/          # Configuration loading
│   ├── db/              # MongoDB client
│   ├── http/            # HTTP handlers and routing
│   ├── models/          # Data models
│   ├── services/        # Business logic (auth, LLM, etc.)
│   └── socket/          # WebSocket manager
├── Dockerfile
├── railway.toml
└── go.mod

Running Tests

go test ./...              # Run all tests
go test -cover ./...       # With coverage
go test ./internal/services # Specific package

Code Quality

go fmt ./...               # Format code
golangci-lint run          # Lint
gosec ./...                # Security scan

Environment Variables Reference

Required

MONGODB_URI=mongodb://...           # MongoDB connection string
MONGODB_DB_NAME=turbostream         # Database name
JWT_SECRET=...                      # Generate with: openssl rand -hex 64
ENCRYPTION_KEY=...                  # Generate with: openssl rand -hex 32
DEFAULT_AI_PROVIDER=anthropic       # Choose your LLM provider
ANTHROPIC_API_KEY=sk-ant-...        # Your provider's API key

Optional

PORT=7210                           # Server port (Railway sets this)
CORS_ORIGIN=http://localhost:3000   # Frontend URL
LLM_MAX_TOKENS=1024                 # Max LLM response tokens
LLM_TEMPERATURE=0.7                 # LLM creativity (0.0-2.0)
LLM_CONTEXT_LIMIT=50                # Max feed entries in context
REQUEST_TIMEOUT_MS=15000            # Request timeout
TOKEN_QUOTA_PER_MONTH=1000000       # Token usage quota

Full configuration guide: turboline.ai/docs/configuration

Resources

API Documentation: turboline.ai/docs/api
Deployment Guide: turboline.ai/docs/deployment
Configuration Guide: turboline.ai/docs/configuration
Community Discussions: GitHub Discussions
Report Issues: GitHub Issues
Video Tutorials: YouTube

License

Licensed under the Mozilla Public License 2.0 (MPL-2.0).

See LICENSE for details.

Contributing

We welcome contributions! Please:

Fork the repository
Create a feature branch from main
Make your changes with clear commit messages
Test thoroughly
Submit a pull request

Contribution guide: turboline.ai/docs/contributing

Made with ❤️ by Turboline AI