Skip to main content
Silo offers a comprehensive suite of features designed to provide a powerful, private, and personalized AI experience.

​
Core Features

  • Open-source and closed-source LLM support
  • Requests proxy in a trusted enclave for closed-source models
  • Private Deep Research (Pro feature)
  • User-sent messages are anonymized before reaching closed-source models
  • E2E-encrypted sync via web app
  • Light and dark mode

​
Privacy Levels

Silo uses a combination of local models, open-source models in Trusted Execution Environments, and closed-source models behind a privacy proxy. See the full Privacy page for details.
  • Requests to closed-source models are proxied through a router running in a TEE, ensuring that model providers cannot match requests to users
  • The embeddings model runs locally on your device
  • Voice models run in a trusted execution environment
  • User-sent messages are anonymized before they are submitted to closed-source models
Main codebase for Silo is open source and available on GitHub. The proxy codebase can be found here and the attestation proxy here.