Protocol for Collective Intelligence

Make Your Knowledge Queryable Without Giving It Up

SyftHub & Syft Space let you create AI-powered endpoints that others can query while your data stays under your control. Share insights, not raw data.

Open SyftHub ->Github

Quickstart
Get up and running in 5 minutes

Installation
Desktop app or Docker deployment

Core Concepts
Learn the building blocks

The Ecosystem

Two Products, One Vision

Together, Syft Space and SyftHub create a decentralized network where knowledge flows without sacrificing control.

Syft Space

Your Personal AI-Powered Knowledge Hub

A decentralized node operated by anyone with knowledge to share. You add your documents, connect an AI model, and create endpoints that others can query. You control who gets access, what they can ask, and whether they pay.

Upload documents or connect vector databases
Connect OpenAI, Anthropic, Ollama, or any provider
Create queryable RAG endpoints
Set access policies and pricing

Deploy a Space ->

SyftHub

The Directory for Collective Intelligence

A registry that maps who has knowledge, what they’re willing to share, and how to reach them. No middleman holding data—just the connective tissue that lets seekers find what they need and providers share on their terms.

Discover knowledge endpoints across the network
Query multiple sources with full attribution
Federated RAG and Mixture of Experts
Transparent source selection

Explore SyftHub ->

The Flow

How It Works

Your Data

AI Model

Endpoint

Your Rules

Controlled Sharing

Core Components

Here goes your text … Select any part of your text to access the formatting toolbar.

Datasets

Upload documents or connect vector databases like Weaviate, Qdrant, or ChromaDB. Files are automatically indexed for AI search.

Models

Connect OpenAI, Anthropic, Ollama, vLLM, or any OpenAI-compatible provider. Your choice, your keys.

Endpoints

Combine datasets and models into queryable RAG endpoints. Configure output as search results, AI summaries, or both.

Policies

Control access, set rate limits, and configure pricing. Your knowledge, your terms.

How it Works

From Your Data to the Network

Two paths, one ecosystem—publish your knowledge or query the collective

Syft Space

Get Queryable in 3 Steps

Publish your knowledge to the network

Add Your Data

Create a dataset from your documents, files, or folders. Automatic processing and indexing begins immediately.

Create an Endpoint

Combine your dataset with an AI model. Configure access policies, pricing, and test with the built-in query interface.

Publish to SyftHub

Make your endpoint discoverable. Choose visibility, monitor usage, and adjust policies anytime.

SyftHub

Query the Network

Pull insights from multiple sources with full attribution

Select Sources

Choose which endpoints to query—specific data owners, categories, or criteria. Selection is explicit and transparent.

Distribute Query

Query goes to all selected endpoints simultaneously. Each processes on its own infrastructure—data never leaves the owner.

Synthesize with Citations

Responses are combined into a coherent answer with inline citations linking each claim to its source.

“Climate change is accelerating faster than predicted^¹ with ice sheet loss contributing 40% more to sea level rise.^²“

^¹ Nature Climate Change – Oxford | ^² Journal of Glaciology – MIT

Applications

What Can You Build?

Publishers & Creators

Make your content AI-queryable for your audience. Let readers ask questions and get answers drawn directly from your work while maintaining attribution.

Research Synthesis

Query multiple research institutions and academic databases to synthesize findings on specific topics with full citations.

Enterprise Intelligence

Combine insights from internal documentation, industry reports, and expert analysis for strategic decisions.

Customer Support

Build an AI assistant that knows your product documentation, answering customer questions accurately without human intervention.

Data Monetization

Share valuable insights from your data without exposing the underlying information. Charge for access on your terms.

Internal Knowledge Base

Turn your team’s scattered documentation into a searchable knowledge hub that anyone on the team can query.

Why SyftHub & Space

Key Benefits

Privacy-First

Your data never leaves your control. Queries are processed on your infrastructure, and only results are shared.

Selective Sharing

Share insights, not raw data. Control exactly what information is accessible through your endpoints.

Monetization

Set your own pricing and access rules. Get paid when your knowledge is used.

Decentralized Network

Part of a growing network while maintaining complete independence. No single point of control.

Build With Syft

Developer Tools

Syft MCP

Model Context Protocol

Connect AI models to Spaces via the Model Context Protocol for seamless integration.

Python SDK

syft-sdk

Query and integrate Spaces in Python applications with full type support.

TypeScript SDK

@syfthub/sdk

Full-featured SDK for Node.js and browser environments.

What if everyone's knowledge made everyone's AI smarter?

A hospital’s research improves a journalist’s questions. A publisher’s archive sharpens a scientist’s model. The best AI would draw on all human knowledge—not just what’s been scraped and frozen into weights, but living expertise that updates, corrects itself, and grows. That’s only possible if contributing knowledge doesn’t mean losing it.

Ready to Make Your Knowledge Queryable?

Join the network of knowledge providers and seekers building the future of AI-powered information sharing.

Open SyftHub ->Deploy a Space Read the Docs

Talk to Our Team

Find out more about OpenMined

This will subscribe you to Product & Protocol News. What to change that later? Use link at the bottom of our emails.

Subscribing confirms you agree with our privacy policy.