Inter-office memo
- To
- Everyone choosing an app to run coding agents
- From
- Mo, Chief Executive
- Date
- November 14, 2026, 4:43 AM CST
- Re
- T3 Code alternatives
T3 Code alternatives in 2026: 8 apps for running coding agents in parallel
The best T3 Code alternative depends on why you’re leaving: Orca sets up the most agents, Paseo has a full phone app and Emdash takes issues from your trackers, all free; Paradev and Conductor add cloud workspaces on Macs; Enjoy runs local models and switches accounts when usage runs out. Stay with T3 Code for a fast, free app on machines you keep running.
Disclosure: Enjoy is our product. We checked every app below against its maker’s own pages on November 14, 2026, and we say where another app is the better pick, T3 Code included.
T3 Code alternatives compared
T3 Code calls itself “the open-source control plane for coding agents.” It’s free, MIT-licensed and popular: its GitHub repository has nearly 60,000 stars, and version 0.1.9 came out on November 11, with nightly builds since. Most reasons to look past it are choices it makes on purpose, such as no local models and no pool of accounts. These are the alternatives, side by side.
| App | Best for | Price, checked November 14, 2026 | Runs on | Agents |
|---|---|---|---|---|
| T3 Code | A fast, free control plane for agents on machines you run | Free | Mac, Windows, Linux | Claude Code, Codex, Cursor, OpenCode, Grok Build, Antigravity, Kiro, Pi, Muse Code, fx, and ACP agents |
| Orca | The most agents set up for you, free and open source | Free | Mac, Windows, Linux | 41 in its docs, and any command-line agent |
| Paseo | A full app on your phone, and agents on several machines | Free; hosted Hub free for 50 runs a month, or $15 a seat a month | Mac, Windows, Linux | Claude Code, Codex, Copilot, OpenCode, Pi, OMP, Antigravity, Muse Code, and ACP agents |
| Emdash | Issues from your trackers, free and open source | Free | Mac, Windows, Linux | 35 in its docs |
| Paradev, formerly Superset | Teams on Macs, with cloud workspaces | Free for one user; Pro $20 a user a month; Team $60; Ultra $200 | Mac; Linux experimental | Claude Code, Codex, OpenCode, and any command-line agent |
| Conductor | A polished Mac app, with cloud sandboxes | No prices listed; cloud workspaces on its Pro, Teams and Enterprise plans | Mac | Claude Code, Codex, Cursor, OpenCode |
| Enjoy | Local models, and accounts that take over when usage runs out | A monthly subscription, free for 7 days, with your phone and one computer (pricing) | Mac, Windows, Linux | Claude Code, Codex, Grok Build, Antigravity, Muse Code, OpenCode |
| Claude Code Desktop | Only Claude, with cloud sessions | Included in Claude Pro ($20 a month), Max ($100), Ultra ($200), Team and Enterprise | Mac, Windows; Linux in beta | Claude Code |
| The ChatGPT desktop app | Only Codex | Included in every ChatGPT plan, from Free; Plus is $20 a month | Mac, Windows; Linux in preview | Codex |
Prices are per month as each maker lists them, before tax. The independent apps run your agents on your own Claude, ChatGPT or other plans and API keys, which those companies bill for separately. Claude Code Desktop and the ChatGPT app come with those plans.
Why people look for a T3 Code alternative
T3 Code does a lot for nothing, and it says plainly what it won’t do. Each reason is in its own words:
- No local models. Its FAQ answers “Does T3 Code support local models?” with “No. T3 Code is a serious developer tool. Locally runnable models are not capable of meaningful engineering work.”
- No pool of accounts. “I don’t think T3 Code is the right place to manage a pile of accounts. If we add an auto-switcher, we are risking TOS bans for our users.” A separate work account is fine, and when a provider reports a reset time, you can choose Resume at reset, but “The environment must be running when the reset arrives.”
- A computer of yours does the work. “The phone is a remote. Your computer does the work and needs to stay on.” There are no hosted workspaces: for remote access, “That machine must stay running and reachable while you work.”
- Its relay is free “for now.” T3 Connect, which links your phone and other devices, “is free too, for now. Every account can connect up to 3 machines running T3 Code.” The FAQ adds: “We don’t want to charge for this, but if the bill gets unacceptable we may have to change course.”
- It sends usage data unless you turn it off. “T3 Code does send basic product usage events (provider, model, token totals, that kind of thing). No prompts, no file contents.” Its data page shows how to stop it with
T3CODE_TELEMETRY_ENABLED=false. - Sharing a thread is a beta. Guest links, which let “a teammate follow one thread in their browser,” are a beta, and each link “works in one browser and expires after 24 hours.”
Orca
Best for: the widest choice of agents set up for you, free.
Orca is “the agent development environment”: “Run Claude Code, Codex, and any other coding agent in parallel, each in its own worktree.” Its README says “Orca is free and open source under the MIT License,” on macOS, Windows and Linux, and its agents page says it “works with any CLI agent,” with 41 set up in its picker, from Claude Code and Codex to Aider, Goose and Kiro.
Its mobile companion for iOS and Android lets you “reply to prompts, review source control, and switch agent accounts,” and its automations “run a prompt on a schedule.” You can switch Claude Code and Codex accounts from the status bar without signing out.
Watch for: the phone is “a remote control for the desktop you already have running,” and for agents that keep going while your laptop sleeps, you run a server yourself: “Orca does not sell managed VPS hosting,” says Ways to run. A new account applies to new sessions only: “Sessions already running keep their original account until restarted.”
Paseo
Best for: a full app on your phone, and agents on several machines.
“Paseo is an open source agentic development environment for desktop, mobile, web, and CLI,” its README says. “Run many agents at once, each in its own worktree, on one machine or several. The mobile app is the full app, native on iOS and Android.” It has “custom implementations for Claude, Codex, OpenCode, Pi, OMP, Antigravity, and Muse Code, and supports many more via ACP,” and its optional relay is end-to-end encrypted: “Paseo can’t read your traffic.” It downloads for macOS, Windows, Linux, iOS and Android.
Price: free, under the Apache 2.0 license. Paseo Hub, which starts agents on your machines from GitHub, Slack and Discord, is free to host yourself; hosted, it’s free for 50 agent runs a month and one seat, or $15 a seat a month for unlimited runs.
Watch for: “It is maintained by one person, full time,” its homepage says, and Hub’s FAQ says “access to the beta stays small and trusted.” Agents run on machines you keep running, as in T3 Code.
Emdash
Best for: sending issues from your trackers straight to agents, free.
“Emdash is a desktop app for running AI coding agents in parallel. Each task runs in its own Git worktree,” its README says. You can send issues from “Linear, GitHub, Jira, GitLab, Asana, Featurebase, Monday.com, Forgejo, Plain, or YouTrack into an agent,” and “work locally or on your own remote machines over SSH/SFTP.” Its docs say “Emdash supports 35 coding agent CLIs,” and its automations “run an agent prompt on a schedule in a project.”
Price: free, under the Apache 2.0 license, on macOS, Windows and Linux.
Watch for: the open-source app is now Emdash Classic, and its homepage says “It still gets fixes, but new development has moved to Emdash 2.0,” which runs agents “on your own infrastructure or Emdash cloud” and sits behind an early-access sign-up with no published price. Schedules need a machine that’s on: “A stopped machine cannot execute a scheduled run.”
Paradev, formerly Superset
Best for: teams on Macs who want cloud workspaces, so agents don’t need a machine of yours.
“Superset is now Paradev,” its homepage says. Its pitch is “Run an Army of Coding Agents,” and “Paradev works with any CLI-based coding agent, including Claude Code, OpenCode, OpenAI Codex, and more.” It says “Paradev is open source under the MIT License,” and an app for iPhone and iPad “lets you check on agents and review their diffs away from your desk.”
Price: the free plan is “For individuals getting started,” with local workspaces, the desktop app and its command line. Pro is $20 a user a month, or $16 billed yearly, and adds remote access, automations and the mobile app. Team is $60, or $54 billed yearly, with cloud workspaces for 150 hours per seat each month. Ultra is $200, with 500 hours, and Enterprise is priced on request.
Watch for: “The desktop app runs on macOS, with an experimental Linux AppImage; Windows is not yet available.” If you use T3 Code on a Windows PC, Paradev isn’t an option there yet.
Conductor
Best for: a polished Mac app, with cloud sandboxes your agents keep working in after you close it.
Conductor is joining Meta, and it called off the shutdown it had announced: its announcement now says “Plans have changed! We’ll still be developing Conductor.” Its homepage leads with “Run a team of coding agents in the cloud.” “Conductor can run Claude Code, Codex, Cursor, and OpenCode sessions in isolated workspaces,” says its harnesses page, and a cloud workspace runs in a sandbox, so your agent “can continue working if you close the Conductor app.” Teammates can share a workspace and “prompt agents together in real-time.”
Price: not listed; its pricing page opens the homepage. Its Cloud FAQ says “Cloud workspaces are available to users on Pro, Teams, or Enterprise plans,” at no extra fee “right now, but we plan to introduce usage-based pricing for cloud compute in the future.” Team plans “are currently invite-only.”
Watch for: “Conductor is not available for Windows or Linux yet,” says its installation page. It runs four agents, without the Antigravity, Grok Build and Kiro that T3 Code runs. And in the cloud, “chat messages from cloud workspaces are stored on Conductor’s servers.”
Enjoy
Best for: local models, accounts that take over when one runs out, and a superagent over all your projects.
Enjoy runs Claude Code, Codex, Grok Build, Antigravity, Muse Code, and local models through OpenCode, on Mac, Windows and Linux, on the AI plans you already have. Each conversation runs its own agent, and an agent can work in a worktree, as it would in T3 Code. What it adds where T3 Code stops:
- Local models. OpenCode in Enjoy works with the providers you set up in it, “including free and local models,” such as one that Ollama or LM Studio runs on your computer.
- Keep going through limits. When an agent runs out of usage, Enjoy switches to another of your accounts that still has usage, or waits for the limit to reset, then continues. Claude Code, Codex and Grok Build can each keep several accounts signed in.
- A superagent over everything. JOY keeps an eye on all your projects: ask it what needs you, catch up on what your agents did, or start work in any project.
- More than code. Projects don’t have to be Git repositories, and workflows can run on a schedule.
- Your phone and any browser. Pair your phone with a code, and use the iPhone app or any browser, end-to-end encrypted.
Like T3 Code, it lets another agent pick up the same conversation: “The new agent gets the conversation so far, including replies from other agents, before it works on your message.”
Price: Enjoy is one monthly subscription, and the first seven days are free, with a card to start. It includes your phone and any browser and one computer; each other computer costs extra each month, and each teammate you invite needs a subscription of their own. See pricing.
Watch for: Enjoy isn’t open source, and it’s a monthly subscription on top of your AI plans once the free week ends, where T3 Code is free. It runs six agents, where T3 Code runs ten and any ACP agent. Your phone and browsers reach your projects only while your computer is awake and running Enjoy, and scheduled workflows run only then too. And T3 Code’s worry about switching accounts is fair for Codex: OpenAI’s terms say you can’t get around its rate limits, and it hasn’t said whether a second paid plan counts. See what each lab says.
Claude Code Desktop
Best for: people who only use Claude, and want work that goes on with the computer off.
Claude Code is the Code tab of Anthropic’s Claude desktop app. You can open new sessions “to work on multiple tasks in parallel,” and cloud sessions “run on Anthropic-managed infrastructure by default and continue even if you close the app or shut down your computer.” You can follow them from the Claude mobile app. It runs on Mac and Windows, with Linux in beta.
Price: “Claude Code is included in all paid plans,” Anthropic’s pricing page says: Pro at $20 a month, or $17 billed yearly, Max at $100, Ultra at $200, Team and Enterprise.
Watch for: it runs only Claude, and “By default, Worktree is off and Claude works in the project folder itself,” so local sessions share one folder unless you turn it on.
The ChatGPT desktop app, for Codex
Best for: people who only use OpenAI.
OpenAI’s desktop app is where Codex runs on your computer: “Run projects in parallel, work with files, use your computer, and keep long-running work moving from one desktop workspace.” You can “Download ChatGPT for macOS, Windows, or Linux,” and on Linux it’s “available in preview.”
Price: on every ChatGPT plan, from Free. In the US, Go is $8 a month and Plus $20, and Pro starts at $100.
Watch for: it runs only Codex, and Free and Go get “GPT-6 Luna at Standard speed in the desktop app, subject to rollout,” not OpenAI’s top models.
When to stay with T3 Code
If you want a fast, free app and you already keep a machine running, stay. It runs ten agents and any ACP agent, with apps for your phone and the web, and its FAQ’s advice is to “Run t3 up on any box and drive it from anywhere.” It has scheduled tasks and webhooks, it resumes a thread when your limit resets, and “There’s no way to pay us to use T3 Code.”
Stay too if Claude’s five-hour limit was what worried you. After reports that apps built on Anthropic’s Agent SDK use the window faster, T3 Code’s FAQ now says “Since v0.1.7, Claude in T3 Code drains your 5h limit at the same rate as Claude Code.”
Move when the gaps above are the ones you hit every day: local models, accounts that run out mid-task, no machine to leave on, or a team to bring along. Most of these apps cost nothing to try, and Enjoy’s first week is free.
How we compared
We read each maker’s homepage, pricing, install and docs pages, and their GitHub repositories, on November 14, 2026, and every one is linked below. We quote their own words for the limits, and we don’t count features a maker hasn’t documented. Roundups at AlternativeTo and OpenAlternative list more apps, including Happy, Jean and Synara, and editors such as Cursor, which we didn’t compare here. This category changes weekly, so we recheck this guide against its sources at least once a month, and the date at the foot of the page says when we last did.
13.Questions.
What’s the best alternative to T3 Code?
- It depends on why you’re leaving. Orca sets up the most agents, Paseo has a full phone app, and Emdash sends issues from your trackers to agents, all free and open source; Paradev and Conductor add cloud workspaces for teams on Macs; Enjoy runs local models and switches accounts when usage runs out; and Claude’s or ChatGPT’s own app will do for one lab.
Is T3 Code free?
- Yes. Its FAQ says “There’s no way to pay us to use T3 Code,” and it’s MIT-licensed. T3 Connect, its relay for your phone and other devices, “is free too, for now,” for up to 3 machines. You still pay for your own Claude, ChatGPT or other plans.
Does T3 Code support local models?
- No. Its FAQ says “Locally runnable models are not capable of meaningful engineering work.” Enjoy runs local models through OpenCode, and Orca, Paseo and Emdash run OpenCode too.
Can T3 Code switch accounts when I run out of usage?
- No, on purpose. Its FAQ says “If we add an auto-switcher, we are risking TOS bans for our users.” It can resume a thread when your limit resets. Enjoy moves the conversation to another of your accounts that still has usage, or waits for the reset; Orca lets you pick another account for new sessions.
What’s a free, open-source alternative to T3 Code?
- Orca, under the MIT license, and Paseo and Emdash, under Apache 2.0. All three are free on Mac, Windows and Linux. Paradev is under the MIT license too, with a free plan for one user, on Mac with an experimental Linux build.
Do I still need a Claude or ChatGPT plan?
- Yes. T3 Code and the independent apps here run agents on your own Claude, ChatGPT or other plans and API keys, which those companies bill for separately. Claude Code Desktop and the ChatGPT desktop app come with those plans.