Skip to content
Skip to content
PilotPods

Compare

How PilotPods compares

A terminal agent, a hosted platform or your own scripts can each be the right choice. Here is where PilotPods is stronger, and where it is not.

  • Runs on your Mac
  • Your engine sign ins
  • Mac only today

Which one fits you

Overview

Three alternatives, side by side

Each card says what the option does well, what PilotPods adds, and when to stay put.

One agent in a terminal

Simple and fast for one task you watch.

What PilotPods adds

  • Many agents at once, grouped in pods led by a Pod COS
  • A daily spend cap and approvals before a risky step
  • State survives a restart, so work is not lost with the window
  • One STOP ALL command for every session

Stay in the terminal if you run one task at a time and watch it.

Hosted agent platforms

No setup, always on, and easy to share with a team.

What PilotPods adds

  • Everything in one SQLite file on your Mac, no hosted service
  • Engines use your own sign ins: claude, codex, antigravity, gemini
  • Zero outbound calls from PilotPods itself, provable with one command
  • No per seat fee for the orchestration layer

Choose hosted if you need a cloud runtime, uptime while your Mac is off, or team sharing.

Home made scripts

Full control and exactly the behaviour you wrote.

What PilotPods adds

  • Durable tasks, heartbeats and an append-only audit log built in
  • Isolated instances with their own data, ports and Keychain prefix
  • Backup, restore, upgrade and doctor commands already written
  • A local web console instead of tailing logs

Keep scripts for a narrow job you fully understand and rarely change.

Side by side

The comparison table

What each approach gives you today. Rows marked Coming soon are not shipped.

Swipe the table sideways to see every column.

CapabilityPilotPodsOne agent in a terminalHosted platformHome made scripts
Runs on your own MacYesYesNo, runs in the provider cloudYes
Many agents at onceYes, in podsOne per windowYesIf you build it
Spend cap and approvalsBuilt inNoVaries by providerIf you build it
Audit logAppend-only, built inTerminal scrollbackVaries by providerIf you build it
Survives a restartYes, SQLite stateNoYesIf you build it
Kill switchstop --allClose the windowVaries by providerIf you build it
Bring your own engine sign inYesYesOften the platform's modelsYes
Always on when your Mac is offNoNoYesNo
Works on Windows or LinuxNo, Mac only todayYesYesYes
Team sharingComing soonNoYesIf you build it

Check it yourself

Claims you can verify with a command

These are real CLI commands. Run them on your own machine.

  • verify-isolation proves zero outbound network calls, no tracking code and no writes outside the data dir.
  • doctor checks service state and isolation between instances.
  • audit lists wake, terminate, approval and spend events.
  • stop --all is the kill switch for every daemon and session in one instance.
Verify, then stay in control
$ ./pilotpods verify-isolation --instance default$ ./pilotpods status --instance default$ ./pilotpods doctor --instance default$ ./pilotpods audit --instance default$ ./pilotpods stop --all --instance default

Questions

Common questions

Short answers, no marketing.

No. It runs engines such as claude, codex, antigravity and gemini and organises them. Your agents still do the work.

Run your first pod in minutes

Install on a Mac, create an instance, and open the console. Everything stays on your machine.

Get started Read the docs
  • MIT licence
  • No telemetry
  • CLI and console
Quick start
$ ./install.sh$ ./pilotpods init --instance default$ ./pilotpods ui --open