Welcome
Olace is your personal AI network. Run AI models on your own computer, reach them from every device you carry, and keep your conversations in sync, encrypted with keys only you hold.
How it fits together
- Host. A machine with a GPU runs your models through Ollama, LM Studio, or llama.cpp. Your computer, a home server, or a cloud VM running the headless daemon; Olace installs and manages the runtimes for you.
- Reach. Any paired device streams from that host over an end-to-end encrypted connection, directly on your local network or through an encrypted relay when you are away. Your phone, a laptop, another computer. No VPN, no port forwarding.
- Continuity. Conversations sync and back up across your devices with zero-knowledge encryption. Olace cannot read them.
- Cloud, when you choose it. An optional credit pool runs zero-data-retention open models for the times your machine is unavailable, or bring your own keys.
What the network gives you
- Remote management. The operational CLI commands take
--device: install and remove models on your host from your laptop, free its GPU, tune its settings, read its logs, restart its service, from wherever you are. - Your other tools come along. The Olace Bridge serves your entire network, local models and every paired host, to any OpenAI-compatible app: coding agents, editors, chat UIs. Your model runs on the big GPU at home; your work happens wherever you are.
- Conversations follow you. Zero-knowledge sync means you close the laptop mid-thought and finish on your phone, and Olace can never read a word of it.
- Useful on day one. Native apps for every platform ship a full workspace (chat, projects, research, web access, images, voice), so remote access never sits idle waiting for you to build something around it.
The pieces
| Piece | What it is |
|---|---|
| The app | Native apps for Windows, macOS, Linux, iPhone, iPad, Android, and the web. |
| The daemon | A small background service for Windows, macOS, and Linux. It keeps local AI and Remote Access available even when the app is closed. |
| The CLI | The olace command. Manage models, pairing, settings, and chat from the terminal, here or on any paired host. |
| Olace Bridge | An OpenAI-compatible endpoint at 127.0.0.1:5578/v1 that gives any other app access to your whole model network. |
Start here
New to Olace? The Quickstart takes you from install to a model in your pocket in a few minutes. After that:
- Remote Access: how pairing works and what it can do.
- Encrypted sync: zero-knowledge continuity across your devices.
- Privacy modes: exactly where your words are allowed to go.