Choose an install
Azhi Flow has two ways to run. They run the same code.
| Team install | Local mode (experimental) | |
|---|---|---|
| For | A team, or anything you want to keep running | One person on one machine |
| Parts | PostgreSQL 16 + pgvector, Temporal and the Azhi server in Docker, plus one or more workers | One process: embedded PostgreSQL (PGlite), Temporal’s dev server, the server and a worker |
| Needs | Git, Node.js 22+, Docker with Compose | Git, Node.js 22+ (Python for script nodes) |
| Command | deploy/install.sh | npx azhi up |
| Data | Docker volumes | ~/.azhi (%USERPROFILE%\.azhi on Windows) |
Platform support
Section titled “Platform support”| Platform | Team install | Local mode |
|---|---|---|
| Linux | Server and workers | Yes |
| Windows 10/11 | Inside WSL2 (Ubuntu) | Yes, native, no WSL |
| macOS | Server in Docker, development worker. Tested and working; workers not certified. | Not yet tested |
Common requirements
Section titled “Common requirements”- Git and Node.js 22 or newer (
enginesinpackage.json). - Docker with the Compose plugin for the team install.
- For script nodes, on each worker host: uv (Python 3.12) and/or Bun. The worker reports what it found, and the run plan marks anything missing as a blocker.
- A model credential: an Anthropic or OpenAI API key, or a Copilot / ChatGPT / Claude plan sign-in. See Use a subscription, not API keys.
Documented hardware (team install)
Section titled “Documented hardware (team install)”| CPU | 4 vCPU (x86_64) |
| Memory | 16 GiB (the three containers use about 420 MiB idle; leave room for load) |
| Disk | 20 GiB free (images take about 1.7 GiB) |
| OS | Linux 6.x with Docker Engine and the Compose plugin |
| Network | Outbound HTTPS to Docker Hub (or a mirror) and the npm registry |
Measured on 2026-10-04: server image build 39 s, start to all three containers healthy 16 s (fresh volumes, migrations included). Image pulls were not measured; at 50 Mbit/s a cold install is estimated at 2 to 3 minutes.
Pick your guide
Section titled “Pick your guide”LinuxServer with deploy/install.sh, then workers.
macOSDocker Desktop for the server, native CLI and development worker.
WindowsWSL2 for the team install, or native local mode.
Local modeOne command, no Docker.