11 Tool Comparison
11.1 Quick reference
All of these tools work through the clinic’s shared gateway. Pick whichever one you’re most comfortable with — or try a few.
| Claude Code | OpenCode | Copilot CLI | Gemini CLI | |
|---|---|---|---|---|
| Start | claude |
opencode |
copilot |
gemini |
| Switch model | /model name |
/models menu |
Quit, change env var, restart | /model name |
| Check usage | /usage |
— | — | — |
| Clear context | /clear |
— | — | /clear |
| Status | /status |
— | — | — |
| Best for | Deep coding tasks, Claude models | Trying open models, flexible config | Quick questions | Google ecosystem |
11.2 Key differences
Claude Code is the most full-featured for this clinic. It has built-in commands for managing context (/clear), checking status (/status, /usage), and switching models (/model). It’s designed around Claude and works well with CLAUDE.md project files.
OpenCode is open-source and has the most flexible model configuration. It works well with both Claude and open models through the gateway. Good if you want to compare models side by side.
GitHub Copilot CLI is straightforward to set up. To switch models you need to quit and restart with a different COPILOT_MODEL environment variable. No GitHub account needed when using the gateway.
Gemini CLI is Google’s tool. It follows the same agentic pattern. Configuration is via environment variables.
11.3 Common commands across tools
Regardless of which tool you use, you’ll do the same basic things:
- Give a prompt — type what you want the agent to do
- Approve tool use — most tools ask permission before editing files or running commands
- Review output — always check what the agent did before moving on
- End the session — close the tool or clear context when switching tasks