QuickStart
Get Jan Agent running and finish a real task in a few minutes. You need a terminal and an API key for one model provider.
Install Nightly
Jan Agent is a preview. The installer on dev pulls from the agent-nightly channel - expect
nightly-quality builds.
curl -fsSL https://delta.jan.ai/jan-cli/install-jan-agent.sh | bash
Installs to ~/.local/bin by default. Override it with JAN_INSTALL_DIR=/usr/local/bin, and make
sure the directory is on your PATH.
Check it worked:
jan --version
Already have the Jan repo checked out? scripts/install-jan-agent.sh --source builds from your
checkout instead of downloading. Source builds don't self-update, because the update channel is
stamped in by CI.
Connect a provider
Jan Agent has no local inference engine - it calls a remote provider. Point it at one:
jan config set --provider anthropic --api-key sk-ant-...
Or sign in to Tokamak, which needs no key of your own:
jan login
Confirm what's configured (keys are redacted):
jan config list
Already using Jan Desktop? Jan Agent picks up the providers you configured there, so you can skip this step entirely.
Open the console
Change into the project you want to work on, then:
jan
You get an interactive console: type a message, press Enter. The agent works in the directory you launched it from.
Give it a task
Start with something small and verifiable:
Read the README and tell me how to run the tests
It will read files and answer. Nothing has changed on disk yet - reading is free.
Now ask for a change:
Add a --version flag to the CLI and update the README to document it
This time the agent asks before it writes:
write src/cli.rs y yes, once a yes, and don't ask again this session n no
Press y to allow it once, a to stop being asked for that kind of call, or n to refuse. You stay in the loop for every write and every shell command.
Watch it work
While the agent runs you'll see each tool call as it happens, and a receipt when the turn ends:
2026-07-29 11:32:02 ↑ 43K ↓ 1.1K ⏱ 1.8s ⚡ 109.8/s
That's the time, context sent, tokens produced, how long it took, and throughput.
Press Esc at any point to cancel the current turn. Press Ctrl+D to quit.
Pick up where you left off
Sessions are saved per project. To continue the most recent one:
jan -c
Or choose from a list inside the console with /resume.
Try these next
Three things worth doing on your first day:
Explore before changing. Start read-only so the agent can investigate without touching anything:
jan --plan
It will study the project and propose a plan. Type /plan exit when you're ready to let it act.
Set a finish line. Instead of babysitting turn by turn, tell it what "done" means:
/goal all tests in tests/auth pass
The agent keeps working and re-checking until that holds.
Seed a task from the shell. Handy in a script or an alias:
jan --task "fix the failing test in tests/auth"
Where to go next
Every slash command and keyboard shortcut.
What the agent can reach, and how to widen or narrow it.
Per-project instructions, models, and budgets.
Split a big job across agents running in parallel.
Troubleshooting
jan: command not found - the install directory isn't on your PATH. The installer prints a
note when it detects this. Add ~/.local/bin to your shell profile.
"no providers configured" - run jan config set --provider <id> --api-key <key>, or jan login
for Tokamak. jan config path prints the file it reads.
The agent won't edit anything - you're in plan mode. Type /plan exit.
Model errors or an unexpected model - /model inside the console lists what's available and
switches on the spot. jan config list shows which providers are wired up.