Unlock the world's distributed intelligence
An open-source protocol for combining private and public models into powerful ensembles - while keeping you in control.
The protocol for Network Sourced AI
Most of the world's intelligence is locked away—in private datasets, proprietary models, local envs, and individual expertise. Centralizing it isn't just hard — it doesn't scale. Intelligence is inherently distributed, and thus, developing AI must be collaborative.
This open-source protocol makes it simple to combine all of these models into powerful, permissioned ensembles.
It respects ownership, rewards contributions, and enables scale without sacrificing control.
Stay in control
Bring your local LLM, RAG or data to the network & control who can see or use it. You keep full ownership over your assets.
Outperform centralised models
Use our unified API Spec to easily ensemble public and private models and share it with others.
Get rewarded
Set your own pricing and control how your assets are used and compensated.
Steps to join the network
Installation
Get Started in Minutes
Install SyftBox and llm-protocol. Initiate your first model project with our CLI helper.
Model Linking
Unified API Spec
Implement our unified API spec to link your model to the network.
Publish
Set Your Terms
Choose your pricing model and set access controls.
Build Ensemble
Reach SOTA
Discover and ensemble various models on the network with yours.
Want to try it out?
Join our early access program and help shape the future of distributed AI