CodebuffAI

freebuff: Use AI Coding Agents Across Your Workflow

Freebuff is a TypeScript project for accessing AI agents across the terminal, desktop, browser, and GitHub repositories. It combines specialized agents with a catalog of models, with free access supported by text ads and subject to regional and capacity limits.

TypeScriptApache-2.0AI AgentsCoding AgentCLI
View on GitHub Compare
Stars
13.1k
Forks
1.4k
Stars, last 30 days
+602

Our take

Health score

56/100

How it is scored
Activity
74
Community
67
Issues
61
Pull requests
20

+602 stars in the last 30 days

Project history

Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.

Selected snapshot

Stars

14,173

Forks

1,521

Change since Nov 14, 2026: +15 stars·+2 forks

Overview

Freebuff brings AI agents for coding, app building, repository work, and research into several interfaces: a CLI, desktop app, web app, hosted GitHub workflow, and chat. Its multi-agent approach can gather project context, plan tasks, edit code, run tools, and review results rather than routing every task through a single prompt.

It is aimed at developers and other users who want agent-assisted work without bringing their own API key. Included model access is supported by text ads, and availability depends on region, access tier, product, and capacity. The repository is a TypeScript monorepo built with Bun and is built on the Codebuff multi-agent framework.

Key Features

  • CLI for working on projects from a terminal.
  • Desktop app for running agents locally, including concurrent agents in separate workspaces.
  • Web app for building and shipping full-stack applications, with hosted sandboxes and previews.
  • Cloud workflows for running agents on GitHub repositories.
  • Chat interface for research and general AI tasks.
  • Specialized agents for codebase context, implementation, review, documentation research, and browser use.
  • A curated model catalog with access and limits that vary by model and user tier.
  • Support in Freebuff Desktop for locally installed Claude Code and Codex agents using existing provider accounts.

Use Cases

  • A developer can use the CLI to ask an agent to find relevant files, make a change, and run project checks.
  • A team can connect a GitHub repository to run hosted agent workflows without setting up a local environment for each task.
  • A solo builder can use the web app to prototype a full-stack application with a sandbox and preview workflow.
  • A developer handling several tasks can use Desktop to run parallel agents in isolated workspaces.
  • A researcher or developer can use Chat and browser-enabled agents to investigate documentation or test an application.

Getting Started

Install and launch the CLI in a project:

npm install -g freebuff
cd ~/my-project
freebuff

See the README for the other products, model access details, and contribution instructions.

Alternatives

  • opencode: OpenCode focuses on coding workflows with plan and build agents, while Freebuff spans multiple interfaces and offers a catalog of specialized agents and models.
  • pi: Pi is an extensible toolkit and CLI for using or adapting agents across LLM providers, rather than a multi-interface service with free, ad-supported access.
  • kilocode: Kilo Code centers on IDE and command-line coding workflows, with model choice and task modes, while Freebuff also targets browser, desktop, and GitHub use.
  • grok-build: Grok Build is a Rust terminal and CI coding agent, whereas Freebuff offers agents across terminal, desktop, browser, and GitHub.
ProjectLanguageLicenseStarsStatus
freebuffTypeScriptApache-2.013.1kNot checked yet
opencodeTypeScriptMIT212kActive
PiTypeScriptMIT117kActive
kilocodeTypeScriptMIT27.5kActive
grok-buildRustApache-2.027.3kNew

Considerations

  • Model availability, limits, and access depend on region, product, account tier, and current capacity. Free access is supported by text ads.
  • The README describes data use and retention conditions, including cases where a model or feature may use submissions for AI training. Review the Privacy Policy before sharing sensitive prompts, code, or repository data.
  • Developing the monorepo locally requires Bun, Docker, and a configured .env.local; the contributing guide contains setup details.
  • The README describes Freebuff as built on Codebuff and links to its SDK and documentation for custom agents and integrations.

Found this useful?

Share it with someone who would like freebuff.

Comparisons