Conductor
Claude Code parallelization — run multiple instances simultaneously
Install with Homebrew
brew install --cask conductorQuick Take: Conductor
Install Conductor on a Mac when you already pay for Claude Code, Codex, or Cursor and keep burning time on sequential agent sessions. Free local worktrees are enough to learn the review habit; Pro at $50/month buys Conductor Cloud (8-core/16GB us-east-1 Vercel sandboxes) plus multiplayer and API; Teams at $60/user/month adds admin and SSO for 5+ seats. The Melty Labs / YC S24 $22M Series A and changelog through 0.79.0 (Cloud Polish, early August 2026) reduce abandonment risk. Pass if you need a native Windows or Linux desktop client, or if cloud transcript storage on Conductor's servers breaks your policy. Local Free workspaces still keep session data on device.
Best For
- •Mac developers running multiple independent agent tasks daily
- •Teams that want Free local orchestration now and Pro/Teams cloud + multiplayer later
- •Shops standardizing on Claude Code and/or Codex and/or Cursor rather than a single vendor UI
What is Conductor?
Conductor (conductor.build) is a Mac desktop app for running parallel coding agents in isolated git worktrees. Unlike older descriptions that framed it as Claude-Code-only, the 2026 product runs Claude Code, OpenAI Codex, and Cursor agents side by side, with experimental OpenCode support. You add a repository, spawn agents into separate worktrees, watch status from a unified dashboard, then review diffs and merge when the work is good. Local parallelism remains free; you bring your own model subscriptions and API keys. Paid tiers unlock Conductor Cloud (Vercel sandboxes), multiplayer, team admin, SSO, and the Conductor API. The problem Conductor solves is sequential agent bottlenecks and messy multi-window setups. One Claude Code session is capable but serial. Manually juggling several terminals, checkouts, and PR tabs does not scale when you have a backlog of independent bugs and features. Conductor makes isolation the default: each workspace is a git worktree, so agents do not overwrite each other's uncommitted files. You see what each agent is doing, steer or interrupt, and collect reviewable branches instead of hoping scrollback tells the truth. Pricing as published on conductor.build/pricing (August 2026): Free at $0 for parallel agents on local Mac workspaces with BYO subscriptions and keys. Pro at $50/month adds Conductor Cloud workspace hours, multiplayer, teams up to 5, the Conductor API, and a mobile app marked as coming soon. Teams at $60/month per user is for organizations of 5+ with admin portal, centralized billing, and SSO. Enterprise is custom with DPA, purchase-order billing, SCIM, custom security controls, SLA, and dedicated support. Model spend is separate, Anthropic Claude Max is commonly recommended when you run heavy Claude Code parallelism; Codex and Cursor costs follow those vendors' plans. Cloud compute is included for now on paid plans, with Conductor noting future usage-based cloud pricing may arrive. Cloud workspaces (Pro/Teams/Enterprise) provision 8-core / 16GB Amazon Linux 2023 sandboxes via Vercel in us-east-1, preloaded with Node.js 24, npm, pnpm, Python 3, git, GitHub CLI, ripgrep, git-lfs, tmux, and vim. Agents keep working after you close the Mac app, unlike local sessions that die when the machine sleeps. Privacy differs: local Conductor sends messages to your model provider and keeps session data on device; cloud sessions store chat inputs and outputs on Conductor's servers because the sandbox is managed. Company context: Conductor is Melty Labs / YC S24, founded by Charlie Holtz and Jackson de Campos, and raised a $22M Series A in 2026. Release release rhythm has been aggressive, changelog entries through 0.79.0 (Cloud Polish, early August 2026) after 0.78.0 introduced Conductor Cloud, with earlier 2026 milestones for Cursor support and Dispatcher, OpenCode, multiplayer/API early access, and current model pickers spanning Claude Opus 5 / Sonnet 5 / Fable 5, GPT-5.6, and Grok 4.5 class stacks. Competitors worth naming: CodexMonitor (open-source Codex-only Tauri command center, maintenance paused), Claude Code alone (same quality, no orchestrator), Cursor background agents (Cursor-native parallelism), and Meta Muse Code (terminal agent narrative rival). Conductor's bet is Mac-native UX plus multi-runtime worktrees, now extended into cloud and multiplayer for teams willing to pay. Honest fit: Solo Mac developers with Claude Max or Codex seats get useful throughput on Free. Agencies and startups that want always-on sandboxes, shared workspaces, and admin controls should evaluate Pro or Teams. Windows/Linux-only shops are out of scope for the desktop app today. Security-conscious teams should read the local-vs-cloud data FAQ carefully before moving secrets into cloud workspaces. Extended guidance for engineering managers: treat Conductor as throughput infrastructure. It does not invent review quality. Parallel agents amplify good tickets and amplify vague ones. Require scoped prompts, acceptance checks, and human review before merge. Budget model spend separately from the $50 Pro seat, four Opus-class agents can burn a Max pool quickly. Prefer local workspaces for proprietary code when policy forbids vendor-stored transcripts; use cloud when laptop lid-close would otherwise kill overnight jobs. Pair with branch protections and CI so agent PRs cannot skip tests. Revisit seat count quarterly as Muse Code, official Codex desktop, and Cursor evolve. Procurement checklist: (1) confirm macOS fleet, (2) map who needs Free vs Pro vs Teams, (3) decide cloud data handling, (4) set Anthropic/OpenAI/Cursor spend caps, (5) standardize worktree cleanup, (6) train on review-before-merge, (7) keep a bake-off against CodexMonitor and vendor-native agents so Conductor stays a deliberate choice.
Closer look: From local worktrees to Conductor Cloud
Conductor's arc tracks the broader agent market: first make parallel local agents safe with git worktrees, then productize always-on compute and multiplayer once usage patterns prove themselves. Free local remains the wedge; Pro cloud is the monetization and retention layer for teams that refuse to keep laptops awake. Operators should budget Anthropic, OpenAI, and Cursor spend separately from the Conductor seat price, enforce branch protections on every agent PR, and document when cloud transcript storage is allowed versus when proprietary repos must stay on local Free workspaces. Treat 0.79.x cloud polish as the August 2026 baseline when comparing Conductor against CodexMonitor or vendor-native background agents.
History & Background
Started as a Mac Claude-centric orchestrator, expanded to Codex and Cursor, added Dispatcher and OpenCode experiments, then launched Conductor Cloud on Vercel sandboxes with multiplayer/API early access, culminating in 0.78-0.79 cloud polish in mid/late summer 2026 after the $22M Series A gave the team runway to ship beyond a local-only indie tool.
How It Works
Local mode: Mac app + git worktrees + agent runtime CLIs/subscriptions with session data staying on device. Cloud mode: managed Amazon Linux sandboxes (8-core/16GB) in us-east-1 with a preinstalled toolchain; session persistence requires Conductor-side storage of chat inputs and outputs. Multiplayer shares those cloud workspaces inside a Conductor org.
Ecosystem & Integrations
Interacts with Anthropic Claude Code, OpenAI Codex, Cursor, GitHub CLI tooling, and Vercel Sandboxes. Competes with CodexMonitor, vendor-native agents, and newer terminal agents like Muse Code. Pricing ladders Free, Pro ($50/mo), Teams ($60/user/mo), and Enterprise so individuals can start without a seat tax.
Future Development
Expect continued cloud polish, mobile client delivery, possible usage-based cloud compute billing, and deeper team controls as Enterprise deals expand. Self-hosted cloud is discussed as an Enterprise interest area rather than a GA feature. Keep watching the changelog past 0.79.0 for multiplayer and API maturity signals.
Key Features
Parallel Claude Code, Codex, Cursor, and OpenCode agents
Launch parallel agents from the Mac app using the Claude Code, Codex, and Cursor tools you already pay for. Conductor bundles or integrates Claude Code and Codex for compatibility and supports Cursor (API key) plus experimental OpenCode. Pick the right agent per task instead of forcing one vendor on every ticket.
Isolated git worktrees
Every workspace is a git worktree so agents edit isolated checkouts of the same repository. Merge conflicts become review problems instead of silent clobbering. Discard a bad run by dropping the worktree without wrecking your main working tree.
Unified dashboard, dispatcher, and review
See agent status, pending reviews, diffs, and PR-oriented workflows in one place. Dispatcher-style routing helps assign work; built-in review surfaces reduce the alt-tab tax of many terminals and IDE windows.
Conductor Cloud (Vercel sandboxes)
On Pro and above, create cloud workspaces that keep running after you quit the Mac app. Each sandbox is an 8-core / 16GB Amazon Linux 2023 machine in us-east-1 via Vercel, with a practical toolchain preinstalled. Bring the same subscriptions/keys pattern as local; understand that cloud session data is stored by Conductor.
Multiplayer, Teams admin, and Conductor API
Paid plans add live collaboration in shared org workspaces, team seats (Pro up to 5; Teams 5+), admin portal, centralized billing, SSO on Teams, and a Conductor API for automation. Mobile client is listed as coming soon on Pro.
Rapid model and toolchain updates
Near-weekly changelog releases track frontier coding models and DX polish, recent lines include Claude Opus 5 / Sonnet 5 / Fable 5 support waves, GPT-5.6, Grok 4.5, OpenCode, Cursor support, and cloud polish through 0.79.0. The app stays current so your orchestrator does not lag the agents it runs.
Who Should Use Conductor?
1Solo Mac developer clearing a sprint backlog
On Free, spins four local worktrees: two Claude Code Max agents on backend bugs, one Codex agent on docs, one Cursor agent on UI polish. Reviews diffs in Conductor and merges green branches before lunch.
2Startup eng lead needing overnight agents
Upgrades to Pro for Conductor Cloud. Kicks off long refactor jobs in us-east-1 sandboxes, closes the laptop, and reviews PRs in the morning. Uses multiplayer so a teammate can steer an agent without SSH archaeology.
3Agency on Teams plan
Five-plus developers share centralized billing and SSO. Admin portal manages seats while each engineer BYOs model keys under policy. Enterprise conversation starts when DPA and SCIM become mandatory for a Fortune client.
How to Install Conductor on Mac
Conductor is a macOS app. Install the desktop build from conductor.build, then connect the agent agent CLIs you already use (Claude Code, Codex, and/or Cursor).
Install agent prerequisites
Install and sign into Claude Code and/or Codex CLI as needed. For Cursor agents, have the required API key ready. Confirm each agent runtime works alone in a terminal or editor before orchestrating.
Download Conductor
Get the Mac app from https://conductor.build/ (current line ~0.79.0 per changelog). Open the app and complete first-run permissions for git and local repositories.
Add a repository workspace
Point Conductor at a git repo, create an isolated worktree workspace, and launch a single low-risk agent task to verify diffs and review UI.
Optional: upgrade for Cloud
If you need sandboxes that survive lid close, multiplayer, or API access, choose Pro ($50/mo) or Teams ($60/mo/user) on https://www.conductor.build/pricing and create a cloud workspace.
Pro Tips
- • Start local on Free to learn worktree review habits before paying for Cloud.
- • Claude Max is the practical recommendation when you plan aggressive Claude Code parallelism.
- • Read the cloud privacy FAQ before putting regulated code in Vercel sandboxes, Conductor stores cloud session transcripts.
Configuration Tips
Scope tasks to minimize worktree collisions
Even with worktrees, overlapping feature scopes create painful merges. Assign agents to different modules or clearly sequenced dependencies.
Separate model budget from seat budget
Free/$50/$60 seats do not include Anthropic or OpenAI tokens. Track Max/API burn separately when running 4-8 parallel agents.
Use cloud for long jobs, local for sensitive trees
Cloud shines when the Mac would sleep mid-run. Keep highest-secrecy repos local so transcripts never land on Conductor servers.
Standardize cleanup
Archive or delete finished worktrees on a release rhythm so disk and git worktree lists stay navigable for the whole team.
Alternatives to Conductor
Conductor competes with single-runtime agents, Codex-focused open-source UIs, and vendor-native background agents.
CodexMonitor
Open-source Tauri Codex-only command center with strong stars but paused maintenance after March 2026. Choose CodexMonitor for free cross-platform Codex depth; choose Conductor for multi-runtime Mac orchestration, cloud sandboxes, and funded product velocity.
Claude Code alone
Same Claude agent quality without parallel worktree orchestration or multi-vendor dashboards. Stay solo for deep single-threaded tasks; use Conductor when backlog items are independent and latency-to-merge matters.
Cursor background agents
Cursor-native parallelism stays inside Cursor's product surface. Conductor is better when you also run Claude Code and Codex and want one Mac review cockpit across agent runtimes.
Meta Muse Code
Emerging terminal agent rival. Muse Code is an agent; Conductor is an orchestrator that can sit above multiple agents. Compare on Mac UX, pricing, and which models your team standardizes on.
Pricing
Free ($0): parallel agents, local Mac workspaces, BYO subscriptions/keys. Pro ($50/mo): Cloud workspaces (Vercel sandboxes), multiplayer, teams up to 5, Conductor API, mobile coming soon. Teams ($60/mo/user): 5+ seats, admin portal, centralized billing, SSO. Enterprise: custom with DPA, PO billing, SCIM, custom security, SLA, dedicated support. Model inference costs are separate. Cloud compute currently included on paid plans with future usage-based cloud pricing planned.
Pros
- ✓True parallel agents across Claude Code, Codex, and Cursor with git worktree isolation
- ✓Local Free tier remains useful, pay only when you need cloud, multiplayer, or team controls
- ✓Conductor Cloud offers 8-core/16GB us-east-1 sandboxes that keep working after you close the laptop
- ✓YC S24 + $22M Series A and near-weekly releases through 0.79.0 reduce abandonment risk
- ✓Teams/Enterprise path with SSO, admin, and DPA for growing organizations
Cons
- ✗Desktop app is Mac-only, no native Windows/Linux Conductor client
- ✗Cloud sessions store inputs/outputs on Conductor servers (privacy model differs from local)
- ✗Model costs scale with parallelism; heavy Claude Code fleets often need Max-class spend
- ✗Agents inherit broad local permissions unless you isolate via dedicated machine/VM practices
Community & Support
Conductor supports users through product docs, changelog, and team/enterprise sales paths on conductor.build. Public research and directory write-ups (Ry Walker, AI IDE lists, Codepick guides) track positioning for Mac agent orchestration. Venture funding and a fast release train mean commercial support channels matter more than GitHub-only indie projects. Teams plan onboarding is invite-oriented, contact Conductor for 5+ seat rollouts.
Frequently Asked Questions about Conductor
About the Author
Expert Tips for Conductor
Users who stick with Conductor cite the worktree mental model, parallel agents feel safe because isolation is default, not a bash ritual.
The Free local tier plus paid cloud is a clean story: try orchestration without a seat tax, then pay when lid-close killing jobs becomes the bottleneck.
Security-minded teams should treat cloud transcript storage as a deliberate trade for always-on sandboxes, not a footnote.
Related Technologies & Concepts
Related Topics
AI coding agent orchestration
Tools that run and review multiple autonomous coding agents in parallel.
Sources & References
Fact-CheckedLast verified: Aug 9, 2026
Key Verified Facts
- Conductor is a Mac app for parallel Claude Code, Codex, and Cursor agents in isolated git worktrees, with experimental OpenCode and latest releases around 0.79.0.[cite-1, cite-3, cite-7]
- Pricing: Free $0 local BYO; Pro $50/mo for cloud/multiplayer/API/teams-up-to-5; Teams $60/mo/user for 5+ with admin/SSO; Enterprise custom.[cite-2]
- Cloud workspaces are 8-core/16GB Amazon Linux Vercel sandboxes in us-east-1; cloud chats are stored by Conductor unlike local sessions.[cite-2]
- Conductor is YC S24 / Melty Labs and raised a $22M Series A in 2026.[cite-4]
- Independent guides position Conductor as a local-first Mac orchestrator competing with single-agent workflows and Codex-focused tools like CodexMonitor.[cite-4, cite-5, cite-6]
- 1Conductor, official site
Accessed Aug 9, 2026
"Product home: run parallel Claude Code, Codex, and Cursor agents in isolated workspaces on your Mac; latest release pointers."
- 2Conductor Pricing
Accessed Aug 9, 2026
"Free, Pro $50/mo, Teams $60/mo/user, Enterprise custom; cloud sandbox specs and privacy FAQ."
- 3Conductor Changelog
Accessed Aug 9, 2026
"Release history including 0.79.0 Cloud Polish and 0.78.0 Conductor Cloud introduction."
- 4Conductor, Ry Walker Research
Accessed Aug 9, 2026
"Independent research on worktrees, multi-agent support, YC S24, and $22M Series A context."
- 5Conductor on AI IDE List
Accessed Aug 9, 2026
"Directory profile situating Conductor among AI coding environments and orchestrators."
- 6Conductor.build intro guide, Codepick
Accessed Aug 9, 2026
"Practical introduction to Conductor's parallel agent and worktree workflow."
- 7Conductor Docs
Accessed Aug 9, 2026
"Product documentation hub for setup, workflows, and feature behavior (fallback to changelog if sections move)."
Research queries: Conductor.build pricing Pro Teams 2026; Conductor Cloud Vercel sandbox; Conductor YC Series A