Guideswisp compared

wisp vs T3 Code

T3 Code is a free control surface for agent harnesses, with phone apps in both stores. wisp sends tasks off and hands you a report every turn. Where each is better, from their own docs.

wisp’s Auto: the harness, model and thinking level it picked for this task, shown before you send

T3 Code is the better pick if you want a native app on your phone that drives agents on a computer you leave on. wisp is for handing work off: every turn ends in a report by default, a named crew keeps working on a schedule, Auto picks the harness, and wisp Cloud runs the agents with nothing to install.

Try wisp CloudFree during the preview

Side by side

T3 Codewisp
In its own wordsAn “agent harness control surface”Launch agents in the background and read what they hand back
AgentsCodex, Claude, Cursor, Grok Build, OpenCode, Antigravity and fx, plus ACP Registry agentsClaude Code, Codex, Cursor, opencode, Antigravity, Grok, Copilot
DesktopWindows, macOS and Linux, and a web appmacOS, Windows, Linux. Builds aren’t signed yet
PhoneiOS and Android apps. “The phone is a remote”: your computer does the work and stays onAny phone browser, a Home Screen web app on iPhone with wisp Cloud, and an Android prototype you build from the repo. No native iPhone app
Where agents runYour desktop, a server or a VPSwisp Cloud: sign in and the agents run on a server. Free during the preview
One prompt on several modelsYes: Shift-click models, each gets its own thread and worktreeYes, Compare: a run per harness, side by side, keep the better one
Click an element, send it to the agentYes, in its desktop browser: picked elements and preview annotations land in the composerYes, Pick element in Preview: its HTML and key styles go in the reply box, without a screenshot
PriceFree: “no way to pay us”. More machines on its relay come with T3 Chat plansFree. wisp Cloud is free during the preview
LicenceMITMIT

From each project’s own README and docs on 7 November 2026. We read them; we didn’t run them. They move fast: check before you choose.

Where wisp is different

  • The report. By default every wisp turn ends with a one-screen report of what changed, with the diff and the tests. T3 Code has Visual replies: ask, and the agent answers with a page.
  • The crew. Named agents with a role, a schedule, a notebook and skills check in and report when something changed. T3 Code has scheduled tasks; wisp’s crew keep a notebook from job to job.
  • Auto. wisp picks the harness, model and thinking level per task from what worked before, and shows the pick before you send.
  • A hosted wisp. T3 Code’s phone needs your computer on. wisp Cloud runs the agents on a server with every lid shut, and you sign in with nothing to install. It’s a preview on one small server.

Where T3 Code is better

  • Phone apps. It has apps for iOS and Android. wisp has no native iPhone app, and its Android app is a prototype you build from the repo.
  • Picking elements. Picked elements and preview annotations reach the agent with your note and a screenshot. wisp’s Pick element adds its HTML and key styles as text, without a screenshot.
  • Voice. Dictate into the composer. wisp’s dictation is on Android only.
  • Git hosts. GitHub, GitLab, Forgejo, Gitea, Bitbucket and Azure DevOps. wisp’s Ship it goes through the GitHub CLI.
  • PR watching. It checks a PR every two minutes and wakes the agent when a check fails. wisp gets there with a CI failed trigger you set up.
  • Community. Far larger than wisp’s, which is one owner and a crew.

Pick T3 Code if

  • Your phone is where you want to drive agents, and a computer of yours is always on.
  • You work on GitLab, Bitbucket or another host besides GitHub.
  • You want to choose each model yourself and fan a prompt out across them.

Pick wisp if

  • You want a report every turn by default, without asking for one.
  • You want agents that keep their own memory and run on a schedule.
  • You don’t have a computer that stays on, and want the agents hosted.

Questions

Is T3 Code free?

Yes. Its FAQ says there’s no way to pay for it, and that its relay has a machine limit that goes up with T3 Chat plans. wisp is free too, and both use your own agent plans.

Does T3 Code run agents in the cloud?

Its FAQ says agents run wherever T3 Code runs: your desktop, a server or a VPS, and the phone is a remote. wisp Cloud is a hosted wisp where the agents run on our server.

Can I run one prompt on several agents?

In both. T3 Code does it with Shift-click in the model picker; wisp does it with Compare.

Does wisp have a phone app?

An Android prototype you build from the repo, a Home Screen web app on iPhone with wisp Cloud, and any phone browser. There’s no native iPhone app.

Try it with the lid shut

Sign in with GitHub and give your own cloud wisp a task. Or run wisp on your own computers: same app, MIT.

or run it on your own computers

macOS · Linux and Windows in beta · Node 20+ · MIT

Runs Claude Code, Codex, Cursor CLI, opencode, Antigravity CLI, Grok Build and Copilot CLI