# The \ for Agentic Coding Free, open-source desktop app for running Claude Code, Codex and Cursor in parallel. Worktree per agent, shared specs, agents that message each other. Canonical: https://traycer.ai/ [The software factory is coming](https://traycer.ai/software-factory): Join the waitlist Traycer is an open-source desktop app. Connect your agents, memories, conversations, and artifacts in one shared workspace. Bring your existing subscriptions, collaborate with your team, and continue work from anywhere. ## TRUSTED BY ENGINEERS AT ![Samsung](https://traycer.ai/home/logo-samsung.svg) ![Microsoft](https://traycer.ai/home/logo-2.svg) ![Cisco](https://traycer.ai/home/logo-3.svg) ![Intel](https://traycer.ai/home/logo-4.svg) ![Honeywell](https://traycer.ai/home/logo-5.svg) ![PayPal](https://traycer.ai/home/logo-6.svg) ![CodeRabbit](https://traycer.ai/home/logo-7.svg) ## Bring your own Subscription Flexibility Use the coding agents you already pay for. Traycer runs Claude Code, Codex, OpenCode, and Cursor side by side in one workspace. Claude, Codex, Gemini, Local runtimes, + custom ![Provider logos for Claude, Codex, Gemini, and more orbiting the Traycer mark](https://traycer.ai/home/feature-byo-stack.webp) ## Agent-2-Agent communication Orchestration Agents can talk to each other like they have walkie-talkies: ask questions, request reviews, and hand off work across separate chats. Agents, worktrees, protocol ![A task fanning out to Cursor, Claude Code, Codex, and OpenCode agents that report results back](https://traycer.ai/home/feature-parallel-agents.webp) ## Shared Context, Any Model Continuity Every Task has its own shared filesystem, artifacts, history, and working context. Switch models mid-chat without losing the thread. Decision, history, Specs, Memory ![Workspace sessions across Claude, Codex, and Cursor beside the shared workspace context](https://traycer.ai/home/feature-persistent-context.webp) ## Built-In Skills Workflow Traycer ships with reusable workflows for planning, debugging, reviewing, debating, documenting, and breaking work into tickets. Plan, Execute, Review, Explain ![The /debate, /plan, /execute, and /review skills stacked on a timeline](https://traycer.ai/home/feature-epic-mode.webp) ## Multiplayer workspaces Workflow Humans and agents work in one shared space. Teammates can inspect tasks, comment on artifacts, and steer work without rebuilding context. People, Context, Sync ![Teammates on different devices and providers sharing synced workspace changes](https://traycer.ai/home/feature-team-cloud.webp) ## Driving productivity at scale - 11,000+ developers using Traycer Desktop - 40,000+ tasks created in Traycer Desktop - 22 coding agents supported, from Claude Code and Codex to Cursor and OpenCode ## What developers are saying > Traycer has a level of understanding and functionality that enables deeper technical applications, something other tools just don't deliver. As a non-technical founder, I appreciate how Traycer tackles exact issues, from directory management to setting up external services making it a game changer for my startup. > > Alan Knudson, Founder of BatteryBuilds > Traycer's User Interface is top notch. Then the plan generation to let you know what it intends to do, which I can correct before it writes the code. I deployed a search feature with filters in my react app with Traycer without writing a single line of code myself and everything just worked seamlessly in first try. > > Chinedu, Fullstack Software Engineer > I've spent multiple years assessing the latest Generative AI companies and building solutions alongside them. Traycer has been one of the most well received tools by our complex application design teams, and I personally use it religiously > > Jeremy Alston, Founder of Agents-Space.com > I have been using multiple tools till now. Claude code, Augment code, Cursor, Windsurf, Cline, Codex CLI, and a few more other tools! For the complex tasks, Traycer is the best tool currently in the market. The multi agent mode is great at making the plan and understanding the context better. Kudos to you and your team! > > Sathwikkuncham, Senior Engineer > Traycer delivers a practical solution for AI-driven coding challenges by breaking down high-level intents into manageable tasks and implementing structured guardrails for quality control. Traycer is the difference between just getting something to work and implementing a deployment to work robustly. > > Christoph, CEO of Athlea > I can say with 100 percent certainty that we would not be here right now without Traycer. On go live day we rebuilt our database, webhooks and APIs and switched payment processors entirely. It would have been impossible to get this much code done this fast without it. Traycer Epic Mode is how we did it in just 6 days, work that took us 5 months the first time. > > Zachary Sura, Founder at IQON > Traycer has unlocked a new level of productivity for me, the ability to have multiple agents plan and execute tasks asynchronously makes me feel more like an Orchestra Conductor than Engineer - what a time to be alive > > Janyk Theron, Head of Engineering ## Choose the perfect plan for yourself Whether you're just starting or looking to increase your productivity, Traycer has a plan that fits your needs. | Plan | Who | Yearly (-20%) | Monthly | Includes | | --- | --- | --- | --- | --- | | Free ([Download](https://traycer.ai/download)) | For individuals | $0 Free forever | $0 Free forever | Works with any coding agent (Claude Code, Codex, Cursor, Opencode, ...); AI planning & task tracking; Unlimited agent sessions; Cloud Sync; Device switch; Remote hosts; Mobile app; No token metering; No inference credits required; Agent to agent communication | | Sync ([Get Sync](https://traycer.ai/login)) | For teams | $16/user/month, Billed yearly | $20/user/month, Billed monthly | All Free features plus; Team collaboration | Optional, pay as you go | Item | Detail | | --- | --- | | Traycer Credits | Run models through Traycer, billed pay as you go | | Cloud sandboxes | Additional cost | ## Got questions? We've got answers. Find answers to common questions about Traycer and its capabilities. ### What is Traycer? Traycer is a free, open-source desktop app for Mac, Windows, and Linux. It is a shared workspace where coding agents run side by side, each in its own worktree, with shared specs and tickets, and agents that message each other. ### Is Traycer open source? Yes. Traycer is fully open source. The code is at [https://github.com/traycerai/traycer](https://github.com/traycerai/traycer). ### Which coding agents does Traycer support? Traycer runs Claude Code, Codex, Cursor, and OpenCode side by side. You bring the subscriptions you already have. ### Is Traycer free? Yes. Traycer is free for individuals, including cloud sync and switching devices. Teams use Sync at $16 per user per month billed yearly, or $20 per user per month billed monthly, to share a task with the whole team. ### What does pay as you go mean? Running models through Traycer is paid with Traycer Credits, billed by use, and cloud sandboxes cost extra. Nothing else is metered on Free. More detail can be found here: [https://docs.traycer.ai/account/pricing](https://docs.traycer.ai/account/pricing) ### Where can I find support or join community? Join the community on Discord at [https://discord.gg/DSpebDf6Gr](https://discord.gg/DSpebDf6Gr), or browse the documentation at [https://docs.traycer.ai](https://docs.traycer.ai). ### When do I need Sync? Individuals stay on Free, including cloud sync and device switching. Teams move to Sync at $16/user/month billed yearly, or $20/user/month billed monthly, to share a task with the whole team.