The Model Changes
Your AI assistant updates overnight. Suddenly, code generation breaks. Your workflow is destroyed. Your team is blocked.
High riskFor developers who build with AI
Build with AI.Own your infrastructure.
Stop depending on external AI. rbee (pronounced "are-bee") gives you an OpenAI-compatible API that runs on ALL your home network hardware—GPUs, Macs, workstations—with zero ongoing costs.
Open source (GPL-3.0)OpenAI-compatible APIWorks with Zed & CursorNo cloud required
Zed & Cursor: drop-in via OpenAI APIZero ongoing costs
Build AI tools without vendor lock-in
The Hidden Risk of AI-Assisted Development
You're building complex codebases with AI assistance. But what happens when your provider changes the rules?
Your Hardware. Your Models. Your Control.
rbee orchestrates AI inference across every device in your home network, turning idle hardware into a private, OpenAI-compatible AI platform.
How It Works
- 1Install rbeeRun one command on Windows, macOS, or Linux.
- 2Add Your Hardwarerbee auto-detects GPUs and CPUs across your network.
- 3Download ModelsPull models from Hugging Face or load local GGUF files.
- 4Start BuildingOpenAI-compatible API. Drop-in replacement for your existing code.
From zero to AI infrastructure in 15 minutes
Core capabilities
Swap in the API, scale across your hardware, route with code, and watch jobs stream in real time.
Built for developers who value independence
Build AI agents with llama-orch-utils
TypeScript utilities for LLM pipelines and agentic workflows.
TypeScript
import { invoke } from '@llama-orch/utils';
const response = await invoke({ prompt: 'Generate a TypeScript function that validates email addresses', model: 'llama-3.1-70b', maxTokens: 500});
console.log(response.text);Works with any OpenAI-compatible client.
Start Free. Scale When Ready.
Run rbee free at home. Add collaboration and governance when your team grows.
Every plan includes the full rbee orchestrator. No feature gates. No artificial limits.
Prices exclude VAT. OSS license applies to Home/Lab.
Trusted by Developers Who Value Independence
👨💻
Alex K.
Solo Developer
“Spent $80/mo on Claude. Now I run Llama-70B on my gaming PC + old workstation. Same quality, $0 cost.”
👩💼
Sarah M.
CTO
“We pooled our team's hardware and cut AI spend from $500/mo to zero. OpenAI-compatible API—no code changes.”
👨🔧
Marcus T.
DevOps Engineer
“Cascading shutdown ends orphaned processes and VRAM leaks. Ctrl+C and everything cleans up.”
Open Source
GitHub
Growing
Community
Multi-GPU
Orchestration
€0
Avg. monthly cost
Stop Depending on AI Providers. Start Building Today.
Join developers who've taken control of their AI infrastructure.
100% open source. No credit card required. Install in 15 minutes.