cc-switch is the de facto control panel for developers running multiple AI coding CLIs simultaneously. Built with Tauri 2 and Rust, it abstracts provider configuration across five major tools — Claude Code, Codex, OpenCode, OpenClaw, and Gemini CLI — into a single desktop app with a system tray quick-switch. With 47K+ GitHub stars, it solves the fragmentation problem that emerges when you stop picking one AI coding tool and start picking the right tool for each task. The v3.13.0 release (April 10, 2026) added quota visibility, OpenAI-compatible model auto-discovery, and a lightweight tray-only mode. An open-source project by Jason Young (@Jason_Young1231), validated by the community as the standard multi-agent configuration layer in the Claude Code ecosystem.
cc-switch addresses the configuration fragmentation that emerges when developers stop using one AI coding tool and start using several. Claude Code, Codex, OpenCode, OpenClaw, and Gemini CLI each maintain their own config files, API keys, MCP server lists, and prompt libraries. Switching a provider means editing four different formats. cc-switch collapses all of that into one desktop app.
The tool sits in your system tray and propagates configuration changes across all supported CLIs simultaneously. Add an MCP server once — it appears in Claude Code and Codex and Gemini CLI automatically. Update an API key — all tools get it at once.
Provider management: 50+ built-in presets covering Anthropic, OpenAI, Google, and major third-party providers (OpenRouter, Together, etc.). The v3.13.0 release added OpenAI-compatible /v1/models auto-discovery, so new models surface automatically without waiting for cc-switch to add explicit support.
Local proxy with failover: Route all AI traffic through a local proxy that handles rate limiting, circuit breaking, and automatic failover between providers. If Claude Code hits a rate limit, the proxy can fall over to Gemini or a local Ollama instance without interrupting your workflow.
Usage tracking: A unified dashboard showing spend, tokens, and request logs across all tools. Useful for understanding the actual cost of your multi-agent setup versus single-tool alternatives.
Skills and MCP sync: The integrated skills panel can auto-scan GitHub repositories for agent skills and install them with one click. Combined with cloud sync, your entire multi-agent configuration stays consistent across machines.
cc-switch was partly motivated by platform risk. When Anthropic changed subscription terms affecting OpenClaw users in April 2026, developers who had committed their configuration to a single tool had to manually migrate. cc-switch users switched providers in one click. The local proxy’s failover logic ensures that no single AI vendor can lock you out of your workflow.
brew install --cask cc-switch # macOSWindows: MSI from GitHub Releases. Linux: DEB, RPM, or AppImage.
After installation, configure providers with API keys, import existing MCP servers, set failover priorities, and sync prompts across tools. Setup typically takes 5 minutes.
Developers running two or more AI coding CLIs who are tired of managing separate configuration files for each. Power users who want provider failover, cost visibility across tools, and a stable configuration layer that survives vendor policy changes. Teams that need consistent AI tool configurations across machines.
Open-source AI pair programming tool that works in your terminal to edit code across your entire repository.
AWS's AI-powered coding assistant that helps developers build, deploy, and optimize applications on AWS with code generation and transformation.
Open-source AI coding harness builder that makes AI coding workflows deterministic and repeatable via YAML-defined DAG workflows.