Upgrade
Update the command line tool, then the server.
pipx upgrade fadenstack
faden upgradeGet started
From a bare server to the first answer.
One Linux server runs Fadenstack and needs no GPU. Your GPU machines join from the console.
$ pipx install fadenstack
$ faden deployuv tool install fadenstack works as well. Add -y to faden deploy for an unattended install.
Four steps
Install, connect, pick, invite.
$ pipx install fadenstack
$ faden deploy
✓ Docker and Compose v2 found
✓ Configuration written # random keys and database passwords
✓ HTTPS with a certificate for this server
✓ Services started
Console https://ai.example.internal
API https://ai.example.internal/v1# the line from Machines → Add a machine
$ curl -fsSLO https://ai.example.internal/api/install/faden-agent.sh \
&& sudo sh faden-agent.sh --join https://ai.example.internal
✓ NVIDIA GPU found
✓ Agent running as a service
… Waiting for approval. Code on this machine: K7Q-4TD
✓ Approved in the console. Joined as gpu-01Search Hugging Face › chat models
an-open-chat-model
✓ Fits your cluster: 2 machines, 2 GPUs
Settings filled in for this model
Offered as team-assistant
[ Host ]# the chat
https://ai.example.internal
# the agents: Chrome, Word, Excel, PowerPoint, Outlook
Server https://ai.example.internal
# your apps: any OpenAI client
base_url "https://ai.example.internal/v1"Requirements
What you need.
The server and the GPU machines are separate computers. A small team can start with one server and one GPU machine.
| Item | The server | Each GPU machine |
|---|---|---|
| Operating system | Linux, x86_64 | Linux, x86_64, or an NVIDIA DGX Spark (aarch64) |
| Software | Docker Engine 24 or later with Compose v2; Python 3.12 or later | Docker or Podman with access to the GPU (for Docker, the NVIDIA Container Toolkit), and systemd |
| GPU | None | An NVIDIA GPU with a working driver |
| CPU and memory | At least 4 cores and 6 GB, or 8 GB with the privacy module, MCP servers and skills. 8 cores and 8 to 12 GB recommended. | As the models you run need |
| Disk | At least 60 GB; 100 GB or more recommended, plus space for models | About 15 GB for the runtime image, plus the models it runs |
| Network | Ports 80 and 443 free | Reaches the server. The server reaches a cluster's head machine on its serving port, and a cluster's machines reach each other. Machines without internet get the runtime and the models from the server. |
After the install
Keeping it running.
Back up and restore
One command each, for the databases and the configuration. Restoring is still being tested.
faden backup
faden restore <dir>The full guide
HTTPS with your own certificate, more machines, clusters, upgrades and backups, step by step.
Prefer to watch first?
A seven-minute video, from a bare server to the first chat and the first API call.