Confidential AI coding agents

Plug a confidential GLM Latest into Claude Codeto protect your IP.

Quickstart

Get started in one prompt

Paste one prompt into Claude Code and it sets itself up on Privatemode, guided by the official docs and reversible at any time.

Claude Code logo

Claude Code

Anthropic's agentic CLI for your codebase. Pick a mode: try Privatemode in a single project first, or make every session confidential right away.

Try pasting this into Claude Code. It sets up this project only.

›Set up Privatemode as the confidential inference backend for Claude Code - in this project only, so I can test it and cleanly undo it.
  1. 1. Get your API key from the dashboard
  2. 2. Paste the prompt into Claude Code
  3. 3. Test it: say "undo" to revert everything

No risk: Only local project settings are modified and a local proxy is created. Your overall setup stays untouched.

Security and privacy first

Don't rely on agreements. Rely on modern technology.

End-to-end encrypted, even during processing

Your prompts and code are encrypted on your device and processed inside a hardware-isolated environment whose memory stays encrypted the whole time. Neither Privatemode nor any cloud provider can read them. Not by policy, but by architecture.

Hardware-enforced isolation

Inference runs inside confidential computing environments (AMD SEV, Intel TDX, Nvidia Confidential Computing). Remote attestation verifies the server before every session.

Never stored, never trained on

Your code is processed in memory only. Nothing is stored, nothing is used for training. The stack is open source with reproducible builds.

Fits into how your team already works.

Run Pi on models larger than your own GPU, with your prompts kept confidential.

Pi coding agent logo

Duo Self-Hosted without running GPUs of your own.

GitLab logo

Agentic coding across your whole repo, without the provider reading it.

claude code logo

Confidential coding in OpenCode: select Privatemode as the provider.

OpenCode logo

GitHub Copilot Chat and Cline on confidential open-weight models.

Visual Studio Code logo

Chat and code search for self-hosted Tabby on a confidential model.

the logo of Tabby ML

Xcode Intelligence on a confidential model instead of ChatGPT.

xcode logo

Need custom integration?

Would you like to use Privatemode as the AI provider in a specific tool that we do not yet support?

Let us know!

n8n logo white
Xcode logo
jan ai logo