Everything rbee can do
as a platform.
Complete capabilities reference for rbee distributed AI orchestration platform. Multi-machine GPU coordination, OpenAI-compatible API, SSH-based deployment, and programmable routing policies.
Core Capabilities
OpenAI-compatible API. Multi-machine GPU orchestration. User-scriptable routing. Real-time job streaming.
Multi-Machine GPU Orchestration
Orchestrate GPUs across multiple machines as one unified pool. Connect gaming PCs, workstations, and servers via SSH into a single distributed AI cluster.
Heterogeneous Hardware Support
Mix NVIDIA CUDA, Apple Metal, and CPU workers in one pool. Use RTX 4090, M2 Ultra, and AMD GPUs together seamlessly for distributed AI workloads.
SSH-Based Deployment
Deploy distributed AI models via SSH. No Kubernetes. No Docker. No agents or daemons. Simple SSH deployment across your infrastructure in 5 minutes.
User-Scriptable Routing with Rhai
Programmable routing policies via embedded Rhai scripts. Custom load balancing, cost optimization, and compliance routing without recompilation.
// rate_limit.rhaifn route_request(req) { let user = req.user_id; let count = get_request_count(user); if count > 100 { return reject("Rate limit exceeded"); } return route_to_pool("default");}Write Rhai scripts to control request routing, load balancing, and data governance.
GDPR-Compliant AI Infrastructure
Built-in GDPR compliance for EU organizations. 7-year audit retention, data sovereignty controls, immutable logs. Self-hosted AI with complete data privacy.
Lifetime Pricing
Self-hosted alternative to cloud AI. Pay once, own forever. No subscriptions.
Self-Hosted Alternative to Cloud AI
Pay once, own forever. Core rbee is GPL-3.0 and MIT—completely free. Premium features (€129-499 lifetime) for businesses needing advanced scheduling, telemetry, and GDPR compliance.
No subscriptions. No recurring fees. No vendor lock-in.
Complete AI Orchestration Platform
Cascading Shutdown
Ctrl+C tears down keeper → queen → hive → workers. No orphans, no VRAM leaks.
Model Catalog
Auto-provision models from Hugging Face with checksum verify and local cache.
Network Orchestration
Run jobs across gaming PCs, workstations, and Macs as one homelab cluster.
CLI & Web UI
Automate with a fast CLI or manage visually in the web UI—your call.
TypeScript SDK
Type-safe utilities for building agents; async/await with full IDE help.
Security First
Six Rust crates: auth, audit logs, input validation, secrets, JWT guardian, and deadlines.
Intelligent Model Management
Automatic model provisioning, caching, and validation. Download once, use everywhere.
Comprehensive Error Handling
Network, resource, model, and process errors handled gracefully. Automatic retries, fallbacks, and recovery.
Real-Time Progress Tracking
SSE-based progress narration. Watch jobs stream in real-time with cancellation support.