Orca

Orca is a desktop application (an "ADE", agent development environment) built for developers who run multiple AI coding agents at once. Instead of juggling terminal tabs for Claude Code, Codex, Cursor, OpenCode, or any other CLI agent, Orca gives each agent its own isolated git worktree, tracks all of them in a single window, and lets you review, annotate, and merge the results. It's aimed at developers who already use one or more coding agents daily and want a structured way to run them side by side, on the desktop or over SSH, with a mobile app to check in remotely.

Key features

• Parallel worktrees: fan one prompt out to several agents, each running in its own isolated git worktree, then compare outputs and merge the winner.
• Broad agent support: works with any CLI agent that runs in a terminal, including Claude Code, Codex, Cursor, GitHub Copilot, OpenCode, Amp, Devin, Goose, Cline, and many others.
• Mobile companion app: monitor running agents, get notified when one finishes, and send follow-up prompts from iOS or Android.
• Built-in terminal: WebGL-rendered terminal with infinite splits and scrollback that survives restarts.
• Design Mode: click a UI element in an embedded Chromium browser to send its HTML, CSS, and a cropped screenshot into an agent's prompt.
• GitHub and Linear integration: browse PRs, issues, and boards inside the app, and open a worktree directly from a task.
• SSH worktrees: run agents on a remote machine with full file editing, git, and terminal access, with auto-reconnect and port forwarding.
• Diff annotation: comment directly on lines of an agent-generated diff and send the feedback back to the agent.
• File drag-and-drop: drag files or images straight into an agent's prompt inside a VS Code-style editor with autosave.
• Orca CLI: script workflows with commands like orca worktree create, snapshot, click, and fill, so agents can drive Orca itself.
• Usage tracking: view Claude and Codex usage, rate-limit resets, and switch between accounts without re-authenticating.

Ideal use cases

Orca fits developers who already run coding agents from the terminal and want to scale that workflow: running the same prompt against several agents to compare quality, keeping long-running agent sessions organized across many branches, or reviewing agent-generated diffs without switching tools. It's a good fit for teams that use GitHub or Linear for task tracking and want agent work tied directly to issues and PRs. The SSH worktree feature makes it useful for anyone who wants agents running on a beefier remote machine instead of a laptop.

It's not the right tool if you only run a single agent occasionally, or if you want a lightweight, terminal-only setup with no desktop GUI. If you need a hosted, no-install SaaS product rather than a local desktop app, this isn't that either. Teams without any git-based workflow or that don't use worktrees will get less value from the core parallel-worktree feature.

Installation

Orca ships as a desktop app for macOS, Windows, and Linux, plus a mobile companion app.

Download a prebuilt binary from onorca.dev/download or grab one directly from GitHub Releases:

# macOS Apple Silicon, macOS Intel, Windows .exe, and Linux AppImage
# are available from the latest GitHub release

Or install via a package manager:

# macOS (Homebrew)
brew install --cask stablyai/orca/orca

# Arch Linux (AUR)
yay -S stably-orca-bin
# or build from source:
yay -S stably-orca-git

For a headless Linux server, run orca serve and follow the headless Linux server guide in the repo's docs.

Mobile companion apps are available on the iOS App Store (or via TestFlight) and as a downloadable Android APK from GitHub Releases. Pair the mobile app with the desktop app to monitor and steer agents remotely.

To contribute or run Orca from source, see the CONTRIBUTING.md guide in the repository.