Conductor Dev Stack
IntermediateRun Claude Code, Codex, and Cursor's agent in parallel on your Mac with Conductor, each in its own workspace.
Published 27 September 2026
About Conductor Dev Stack
Conductor (conductor.build) is a Mac app for running several coding agents in parallel on one repository. Each new workspace creates a git worktree on a fresh branch behind the scenes and starts an agent inside it with your existing login, so every agent gets its own copy of the code and its own terminal, with no way to overwrite another's work.
In this stack agents and models are separate choices. Claude Code is the default agent, with Codex, Cursor's agent, and OpenCode as the alternatives, in any combination. Each agent has a home model family, but Conductor's provider settings widen it: Claude Code can run through OpenRouter, Bedrock, or GLM, OpenCode reaches nearly any provider through OpenRouter or Cerebras, and Cursor's agent brings its own multi-provider menu.
Review happens inside the app. A visual diff viewer shows each workspace's changes, and the GitHub integration turns a finished workspace into a pull request in one click, then lets the agent read and answer review comments. Linear integration links issues to the work that closes them. GitHub sits in the core here rather than as a swappable choice, because the pull-request loop is built around it.
Workspaces run on your Mac by default and stop when it sleeps. Conductor stands apart from other agentic development environments by offering a hosted cloud instead of running on a server of your own: on the Pro plan ($50/month) and above, cloud workspaces run in isolated Linux sandboxes that Conductor hosts, so agents keep working with your Mac closed or off while the Mac app stays the interface. A cloud workspace sleeps after four hours without activity, and Conductor itself can't be installed on your own server. GitHub Actions for testing every agent branch and Docker for per-workspace services are the common additions.
Key Features
- ✓One-click workspaces that each run on their own git worktree and branch
- ✓Pick one or more agents: Claude Code, Codex, Cursor's agent, and OpenCode, side by side
- ✓Provider settings that run Claude Code, Codex, and OpenCode against more than their home models
- ✓Visual diff viewer for reviewing each workspace's changes before merging
- ✓One-click pull requests, with agents reading and answering GitHub review comments
- ✓Cloud workspaces in isolated Linux sandboxes that keep running with your Mac closed (Pro)
- ✓Linear integration linking issues to the workspaces that resolve them
When to Use Conductor Dev Stack
- →Mac developers who want parallel agents without learning git worktrees themselves
- →Giving Claude Code and Codex the same task and opening a pull request from the better result
- →Clearing a backlog of small, independent issues with one agent per issue
- →Teams that review agent work as ordinary GitHub pull requests with comments
- →Keeping agents running in the cloud while away from the desk, on the Pro plan
Pros
- The gentlest learning curve among agent orchestrators: worktrees are created and cleaned up for you
- Free core tier with no per-seat cost, using the agent subscriptions you already have
- Tight GitHub loop from workspace to pull request to answered review comments
- Paid plans add hosted cloud workspaces without setting up a server yourself
Cons
- macOS only, with no Windows or Linux desktop app
- Closed source, so it can't be self-hosted on your own server
- Four supported agents, fewer than cross-platform alternatives like Orca
- Cloud workspaces need the $50/month Pro plan, sleep when idle, and there's no mobile app yet
Coding Agent Options for Conductor Dev Stack
Conductor Dev Stack with Claude Code
Claude Code is the default agent here and the one most Conductor workspaces start with: Anthropic's CLI, strong at multi-file changes. It bills through a Claude plan from $20/month or per token, and Conductor's provider settings can also route it through OpenRouter, Bedrock, or GLM.
Conductor Dev Stack with Codex
OpenAI's agent CLI, included with every ChatGPT plan down to the free tier. A second workspace running Codex on the same task is the quickest way to compare two vendors and keep the better branch. In Conductor it runs through an OpenAI account or Vercel AI Gateway.
Conductor Dev Stack with Cursor CLI
Cursor's agent inside a Conductor workspace, for developers already on a Cursor plan ($20/month and up). It brings Cursor's own model menu, Grok by default with Claude, GPT, and Gemini available, billed through Cursor rather than separate provider keys.
Conductor Dev Stack with OpenCode
Open-source and model-agnostic, OpenCode is the agent that opens the model layer widest in Conductor: through OpenRouter, Baseten, Cerebras, or Vercel AI Gateway it reaches Gemini, DeepSeek, Kimi, and many more, billed per token. A good pick for cheap, high-volume work next to a premium agent.
These are highlighted picks. To see all the tools, check the AI Coding Agents category.
LLM Options for Conductor Dev Stack
Conductor Dev Stack powered by Claude
Claude is the default model, the one Claude Code is built for and a common choice for the workspace doing the hardest task. Cursor's agent and OpenCode can use Claude models too, so two workspaces on different agents can still share one model family.
Conductor Dev Stack powered by OpenAI
OpenAI's GPT models are Codex's home family and are also available through Cursor's agent and OpenCode. Picking them alongside Claude gives a second vendor's approach to the same task, and a ChatGPT plan covers Codex usage without per-token billing.
Conductor Dev Stack powered by Google Gemini
Gemini's long context window suits repository-wide questions and large refactors, and its Flash models are fast and cheap for routine work. In this stack it's reached through Cursor's agent or OpenCode, since neither Claude Code nor Codex runs it.
These are highlighted picks. To see all the tools, check the LLM category.
Conductor Dev Stack Add-ons
Each addition below extends this stack with a capability the base stack works fine without. None are required: include the ones your product actually needs when building this stack, and skip the rest.
Code Review Add-ons
Add code review when you want an AI reading every pull request before it merges: it comments on the diff so bugs, security issues, and inconsistencies surface before a human has to catch them.
Conductor Dev Stack with CodeRabbit
Add CodeRabbit when several agents are opening pull requests faster than one person can read them: it reviews every GitHub or GitLab diff with a summary and line-level comments, with a free tier for open-source repos.
Conductor Dev Stack with Greptile
Add Greptile when parallel agents produce more diffs than one person can review: it indexes the whole repository so its comments carry context beyond the changed files, and it can be self-hosted for full control.
Conductor Dev Stack with Cursor Bugbot
Add Cursor Bugbot when a bug-focused reviewer is enough: it reviews GitHub or GitLab pull requests whatever wrote the code, though it bills through a Cursor account and plan rather than its own subscription.
CI/CD Add-ons
Add CI/CD when you want a dedicated pipeline for running tests, linting, or multi-stage builds before a deploy goes out. Many hosting platforms already redeploy automatically on every push on their own — a CI/CD tool adds the most value on top of that by gating the deploy on a passing test suite, and matters even more when the hosting choice does not auto-deploy at all, such as a self-hosted server.
Conductor Dev Stack with GitHub Actions
Add GitHub Actions when several agents are opening pull requests at once and every branch should be tested automatically before you spend time reviewing it.
These are highlighted picks. To see all the tools, check the CI/CD Pipelines category.
Containerization Add-ons
Add containerization when you want the app packaged the same way across local development, staging, and production, or need to deploy somewhere that isn't a managed serverless platform.
Conductor Dev Stack with Docker
Add Docker when each agent's workspace needs its own database or dev server: a Compose file per workspace keeps ports and data from colliding when several agents run the app at once in Conductor.
These are highlighted picks. To see all the tools, check the Containerization category.
Frequently Asked Questions about Conductor Dev Stack
Can agents keep working when my Mac is closed?
Yes, in cloud workspaces, which come with the $50/month Pro plan and above. They run in isolated Linux sandboxes on Conductor's side, so agents keep going with your Mac closed or off, and the Mac app stays the interface when you come back; you can also open a cloud workspace over SSH or sync it to a local folder. A workspace sleeps after four hours without activity, and a sandbox stops after about 24 hours, so very long runs need checking in. Local workspaces stop when your Mac sleeps, Conductor can't be self-hosted on your own server, and its mobile app is announced but not yet released.
Is Conductor's free tier enough?
For one developer working locally, usually yes: the free tier runs multiple agents in parallel with local workspaces and your own subscriptions or API keys. Pro at $50/month becomes worth it when agents should keep working with the Mac closed, in cloud workspaces, or when up to five people need to work in the same workspaces.
Can I use models other than Claude and OpenAI?
Yes, in several ways. OpenCode reaches Gemini, DeepSeek, Kimi, and many others through OpenRouter, Baseten, Cerebras, or Vercel AI Gateway. Cursor's agent brings its own menu of Grok, Claude, GPT, and Gemini models on a Cursor plan. Conductor's provider settings can also route Claude Code through OpenRouter, Bedrock, or GLM, though Anthropic itself supports only Claude models. Codex runs through OpenAI or Vercel AI Gateway.
Does Conductor work on Windows or Linux?
No, Conductor is a macOS app only. The cloud workspaces on the Pro plan run on Conductor's servers, but you still start and manage them from the Mac app. On Windows or Linux, Orca and T3 Code cover the same parallel-worktree workflow.
How is Conductor different from Orca?
Both give each agent its own git worktree. Conductor is the simpler, more polished Mac app with four agents and a paid hosted cloud. Orca is free and open source, runs on macOS, Windows, and Linux with mobile apps, supports more than 25 agent CLIs, and can self-host on your own server instead of paying for cloud workspaces.
Stacks Related to Conductor Dev Stack
Paseo Dev Stack
Developer
Run Claude Code, Codex, and other coding agents on your own machine with Paseo and steer them from any device.
Claude Code Dev Stack
DeveloperClaude Code in the terminal, desktop app, or web, paired with Claude and the add-ons to run it anywhere.
T3 Code Dev Stack
Developer
One open-source app for Claude Code, Codex, OpenCode, and other coding agents, on desktop and mobile.
Orca Dev Stack
Developer
Run Claude Code, Codex, and other coding agents side by side in Orca, each on its own branch and model.
Scores
Popularity3/5
Conductor is one of the best-known tools for running coding agents in parallel, but that practice is still young and mostly limited to AI-heavy teams.
Learning Curve2/5
Workspaces, branches, and worktrees are created with one click, and review and pull requests happen in a familiar GUI, so running parallel agents needs little more than knowing the agents themselves.
Flexibility3/5
Three agents can run in any combination and Cursor's agent opens up other providers' models, but the app is closed source, macOS-only, and built around GitHub.
Performance4/5
Parallel workspaces finish independent tasks in the time one agent would take for a single task, and the native Mac app stays responsive across many of them.
Portability2/5
The app is closed source and macOS-only, and the pull-request workflow depends on GitHub, though the agents and the code itself carry over to any other setup unchanged.
Tools in the Conductor Dev Stack Stack
Add-ons (optional — add any, or none)
Code Review
CI/CD
Containerization
Conductor Dev Stack Pricing
Conductor's core tier is free, so for a solo developer working locally the bill is the agents: Claude Code starts at $20/month with no free tier, Codex has a limited free tier with Plus at $20/month, both reach $200/month, and Cursor is $20/month. Conductor Pro at $50/month adds cloud workspaces and multiplayer for up to five people; Teams is $60 per user per month. GitHub, GitHub Actions minutes, and Docker Desktop are free for personal use.
Free for local workspaces; Pro is $50/mo with cloud workspaces and multiplayer; Teams is $60/user/mo.
Claude Code from $20/mo (no free tier); Codex free tier, Plus $20/mo; both up to $200/mo; Cursor $20/mo. One subscription per agent.
Free for individual developers; GitHub Team is $4/mo per user for more seats and controls.
GitHub Actions includes free monthly minutes before paid usage kicks in.
Docker Desktop is free for personal use; Docker Pro adds team features from $5/mo.