Build apps and automate your work.
Codewhale is an open-source agent that writes code, runs commands and works with the tools you connect. Use your existing model APIs, or run models locally and on your own servers.
curl -fsSL https://codewhale.net/install.sh | sh
▎ Local demo: add --json to export.py, run tests, and delegate a review. │╭ ✓ • tool done · 3/3 · 100% ││ ▏ checklist: todo_write ││ ▏ ☑ Inspect the export command ││ ▏ ☑ Add the --json flag │╰ ▏ ☑ Run tests and review the change │ │● JSON export is ready. │▏ │▏ Added --json while preserving CSV output. All 3 local tests passed, and the delegated review │▏ found no issues. ┃▏ ┃▏ Try: python3 export.py --json ┃▏ ┃▏ This session uses the local website-demo provider. ┃ ┃Note Workbar is off — no panel renders (/workbar bottom|top|left|right to show it) ┃╭──────────────────────────────────────────────────────────────────────────────────────────────────╮│❯ Type a message… ││ [·] │╰──────────────────────────────────────────────────────────────────────────────────────────────────╯ ● Full Access Shift+Tab to change work (Tab) Work bar (Ctrl+]) ← for agents · ↓ to manage website-demo · website-demo context 7%
What you can do
Describe what you want to build or automate. Codewhale can edit files, run commands and check the result, with access you control.
- Build apps and tools
- Create an app, add a feature or write a script. Codewhale can work through the project files, run the code and test what it builds.
- Automate the work you repeat
- Run workflows from your terminal, scripts or CI. For larger tasks, delegate parts of the work to a Fleet of agents with different models.
- Connect the tools you use
- Add tools through plugins and MCP servers, or use APIs from your own scripts. Each service needs its own setup and authentication.
Use the models you choose
Connect your provider accounts, an OpenAI-compatible endpoint, or local and self-hosted models. Pick a model for the session and for each agent in a Fleet.
Browse models and providers- Your API accounts
- Connect providers such as OpenAI, Anthropic, Google or DeepSeek with your own keys.
- Your gateway
- Use an OpenAI-compatible endpoint and choose the models it serves.
- Your inference
- Run local or self-hosted models with Ollama, vLLM or SGLang.
Work across files and tools
Explore tools and integrations- Files and terminal
- Create files, run commands, inspect data and test what you build. You set the working folder and permissions.
- Plugins and connected apps
- Add skills and tools through plugins and MCP. Review and enable the connections you want the agent to use.
- Browser and computer · preview
- Use browser tools and the Computer Use plugin for work in apps and websites, with the access you grant.
- Sessions you can return to
- Keep the conversation, tool results and work history together. Resume the task in the terminal or its local browser client.
- Teams of agents
- Use Fleet to split a larger job between agents with different roles and models, and follow their progress in one place.
Get started
Install Codewhale, connect a model and open a project folder. You can add plugins and more agents as you need them.
Install Codewhale
Install the latest release on macOS or Linux. Follow any PATH instruction the installer prints, then open a folder where you want to work. The guide covers Windows and other install options.
Full install guidecurl -fsSL https://codewhale.net/install.sh | shConnect your model
Use a model API you already have, or connect local or self-hosted inference. The command below saves a DeepSeek key; local Ollama models do not need an API key.
Connect a providercodewhale auth set --provider deepseekDescribe your first task
Start Codewhale and describe the result: a small app, a script that saves you time, or a report from your data. Choose its permissions, review what it does and steer it as you go.
Set modes and approvalscodewhaleUse multiple agents
Split a larger job across agents with different roles and models. Inside Codewhale, save your team with /fleet setup, then check its runs with /fleet status.
Set up a Fleet/fleet setup /fleet status
Available now and in development
The terminal and local browser client are available now. A native desktop app and a rebuilt hosted web app are in development.
Create an account- Terminal and local browser
- ReleasedInstall on Linux, macOS, or Windows, then run codewhale, or codewhale web for the local browser client. npm and Cargo also work; Android on Termux is a preview.
- CodeWhale GUI (VS Code)
- AvailableA separate, community-maintained project: chat, threads, and file changes in a VS Code sidebar over the same Codewhale Runtime. Install it from the VS Code Marketplace.
- Hosted web app
- Development previewBeing rebuilt to match the desktop app. Today you can sign in, then type /rc in a running terminal session to continue it on the web; hosted task execution is still being qualified.
- Desktop
- Development buildThe native app becoming the main Codewhale client: folders, conversations, and model connections in one window. No public download yet.
- Cloud computers
- In developmentHosted computers that run your tasks.
The terminal, local browser, and GUI need no Codewhale account. Hosted web and desktop use an account. When you use your own provider key, your provider bills that usage.
Contribute to Codewhale
Report a bug, improve the docs or contribute code on GitHub. You can also build plugins and share workflows with other users.
Install on macOS or Linux
GitHub Releases (Binaries) · npm · Cargo · Docker · Windows · Android / Termux · China mirrors
Read the install guidecurl -fsSL https://codewhale.net/install.sh | sh