Skip to main content
📖 The AI Tool Bible
Codewhale logo

Codewhale

Not yet source-checked

Build and automate with the models you choose

Machine-drafted from the vendor's site · not yet source-checkedProfile updated Link checked How we rate

FreeAgents

In short

Codewhale is an open-source terminal agent that edits code and runs tests using your choice of hosted or local AI models. It supports plugins and multi-agent workflows for developers.

Best for

Developers who want to build and automate tasks using open-source agents with their choice of hosted or local AI models.

Skip if

Users who require a fully hosted web app or desktop application, as these are currently in development or preview.

Codewhale is an open-source agent that reads files, edits code, runs commands, and runs tests to check its work. It operates in a terminal or local browser using hosted or local AI models. Users choose the tools and permissions, and the session keeps the conversation and tool results. The tool allows users to ask for concrete results like fixing bugs or understanding projects, and can turn repeated tasks into workflows or scripts. It supports plugins and MCP for connecting additional tools, and includes a 'Fleet' feature to assign parts of a task to agents with different models and roles. The terminal and local browser clients are released for macOS, Linux, and Windows, while hosted web and desktop apps are in development.

Editor's take

The page documents Codewhale as an open-source agent for coding and automation that runs in a terminal or local browser. It highlights features like file editing, command execution, test running, and the ability to use different models via API keys or local instances. The page notes that while the terminal and local browser clients are released, hosted web and desktop versions are still in development. It does not provide specific pricing tiers, as the software is open-source and model costs are billed by the provider.

— The AI Tool Bible editorial team

Pros

  • ✅ Open-source agent
  • ✅ Reads files, edits code, runs commands, and runs tests
  • ✅ Use it in your terminal or local browser
  • ✅ Choose the provider and model for each session
  • ✅ Turn a repeated task into a script or saved workflow
  • ✅ Set permissions before work starts

Cons

  • ⚠️ Hosted web app is a development preview
  • ⚠️ Desktop is a development build
  • ⚠️ Cloud computers are in development
  • ⚠️ Hosted model usage with your own key is billed by that provider

Facts

Platforms
cli, macos, linux, windows, android
API available
Yes
Open source
Yes
Earliest Wayback capture
2021-11-28 (earliest archived capture of the site, not a launch date)

Where this tool sits

Use cases

Understand a projectTurn a repeated task into a workflowInspect a project and make a changeRun testsAssign parts of a task to agents with different models

Frequently asked

How much does Codewhale cost?
Codewhale is free. The provided documentation does not list specific pricing tiers or additional costs for using the open-source agent.
Which AI models can I use with Codewhale?
You can use either hosted or local AI models. The tool is designed to let you choose the specific models you want to build and automate tasks with.
Does Codewhale integrate with other tools?
Yes, it supports plugins and MCP (Model Context Protocol) for connecting additional tools. This allows you to extend its capabilities beyond basic file reading and command execution.
Is there a web or desktop app available?
Hosted web and desktop apps are currently in development or preview. The released clients are for terminal and local browser use on macOS, Linux, and Windows.
Can I assign different tasks to different agents?
Yes, the 'Fleet' feature allows you to assign parts of a task to agents with different models and roles, enabling complex multi-agent workflows.
See alternatives to Codewhale →

Explore related

Reviews