Keep Jan.
Give your models your machines.

AI Commander is not an alternative to Jan, Claude Code, Codex, ChatGPT, or any other AI client. It is the remote-machine harness those tools plug into over MCP: outbound-only shell, detached jobs that outlive the chat, and file transfer on hardware you already own.

Positioning in one line: Jan stays your offline-first chat and model runner; AI Commander is the hands on the box — servers, GPU rigs, NAS, laptops — with no inbound ports and no exposed SSH.

Sibling harness pages: Msty Studio · Chatbox · LM Studio · AnythingLLM · TypingMind · Open WebUI · Claude Code · Codex · Claude Desktop — and the rest on Connect your AI client.

Connect Jan (Settings → MCP Servers)

Jan is an MCP host that launches MCP servers as local commands (Jan docs: MCP). Add AI Commander through its stdio bridge, which talks to the hosted endpoint for you. You need Node.js 20 or later on the machine running Jan.

Open Settings → MCP Servers and click the + in the MCP box.
Server Name aicommander, Command npx, Arguments -y @aicommander/mcp.
Environment variable (optional): AICOMMANDER_TOKEN = your API key.
Check that the server shows as active, then pick a model with tool calling enabled.

Jan → Settings → MCP Servers → +

Server Name:  aicommander
Command:      npx
Arguments:    -y @aicommander/mcp
Env:          AICOMMANDER_TOKEN=<your-api-key>   (optional)

Create an API key in the AI Commander dashboard (see Accounts & API keys) so the agent can list your machines and reach them by alias. No account yet? Leave AICOMMANDER_TOKEN out: the bridge then connects anonymously for a code-only first hour, reaching machines by their AIC-… session code.

Get the details right

Tool approvals: Jan asks you to approve each tool call and shows the exact arguments first. Keep that on while you get used to it; Allow All MCP Tool Permissions pre-approves everything.
Local models: enable tool calling in the model’s capabilities — many open models are weak at tool use, so try a stronger one if the agent ignores the tools.
Long work: start detached jobs (remote_job_start, then remote_job_status / remote_job_logs) instead of one long tool call.
Windows: if npx fails to start, use npx.cmd as the command.

By default, API keys and OAuth connectors only act while the account owner has opened the AI Commander dashboard in the last 24 hours (accounts can disable that gate under Credential security) — see Connect your AI client. AI Commander’s endpoint is first-party and only acts on machines where you installed the agent yourself. Menus and options can differ between Jan versions — check the linked Jan docs for your build.

What you get that the chat alone cannot

Remote shell on Linux, macOS, and Windows — by session code or saved alias.
Detached jobs for builds, test suites, and training on the box that has the repo.
Fleet list + GPU facts so your agent picks the right machine before it burns an hour.
File courier (Pro) for checkpoints and diffs up to 100 MiB.
Outbound-only agent on each machine — no inbound ports, no exposed SSH, no VPN to set up.

Not a replacement — a multiplier

You already use	AI Commander adds
Jan	Settings → MCP Servers → + → `npx -y @aicommander/mcp`
Claude Code	`claude mcp add --transport http …` → remote exec + jobs
Codex / ChatGPT	Same MCP endpoint or REST `/api/v1/exec`

FAQ

Is AI Commander an alternative to Jan?

No. AI Commander is NOT an alternative to Jan. It is a remote-machine harness. You keep Jan and your models (or Claude Code, Codex, ChatGPT, Cursor, and the other clients on /howto/) as your AI front end. AI Commander gives those tools secure remote shell, detached jobs, and file transfer on machines you own — over MCP or REST. It is also not an alternative to Claude Code, Codex, or ChatGPT; it works with them the same way.

How do I connect Jan to AI Commander?

Open Settings → MCP Servers, click +, and add a server named aicommander with command npx and arguments -y @aicommander/mcp. Optionally set the environment variable AICOMMANDER_TOKEN to your AI Commander API key. Jan needs Node.js 20 or later for the bridge; without a key it connects anonymously for a code-only first hour.

Does AI Commander work with local models in Jan?

Yes, as long as the model can call tools: enable tool calling in the model's capabilities in Jan. The model runs where Jan runs; AI Commander only gives it tools that act on your remote machines.

Next steps

Install the agent on a machine you control, add AI Commander under Jan’s Settings → MCP Servers (npx -y @aicommander/mcp), then ask your model for a real task on a remote box.

Install the agent Setup guides Remote coding agents

Keep Jan.Give your models your machines.