Amp vs Claude Code

Overview

Amp is a coding agent and development environment built for frontier models. Instead of keeping one chat in one editor, Amp gives every agent its own orb — an isolated environment that keeps working after you close your laptop. You can start a task, walk away, and check on it from the web or your phone while it continues in the background.

The product is aimed at developers who run several agents at once and want to stop juggling worktrees, terminals and browser tabs. Amp describes the shift as "juggle less, ship more": you direct a team of agents, share threads with teammates, and pick work back up from any device.

What Amp provides

• Orbs — always-on environments for agents. Each orb can run commands, edit code and keep state, and agents inside them continue after your local machine disconnects.
• Remote control — run amp --no-tui on servers, CI runners or any shell, then watch and steer those agents from the web UI or your phone.
• Threads and spaces — long-running conversations grouped into spaces so multiple agents and teammates can work on the same project without stepping on each other.
• Automations — agents can schedule themselves to wake up, check production metrics, compare canary results and report back without a human prompt.
• Production access with OIDC — orbs can be granted (and revoked) access to production systems such as Google Cloud, so an agent can read logs or verify a rollout.
• Team features — service accounts for threads that belong to the team rather than one person, plus sharing so colleagues can follow or take over work.
• Git hosting integrations — connect github.com, gitlab.com or a self-hosted GitLab instance; Amp can also host a repository for a new project and clone it with amp clone.
• Portals — open what an agent built on a remote runner directly from your browser, even when the code is not on your machine.
• Boxes — always-on computers for the things you build with Amp.

How people use it

Typical workflows include handing off a bug report and letting the agent reproduce, patch and test it; running performance rollouts where an agent watches dashboards and confirms a fleet-wide win; reviewing a diff and asking follow-up questions in the same thread; and letting an agent fix its own tooling or documentation when it hits a bad error message.

Amp is model-agnostic in practice — the agent can be pointed at different frontier models — and it exposes the same threads across the desktop app, CLI, web and mobile, so a task can start on one surface and finish on another.

Getting started

1. Create an account at ampcode.com and install the CLI or desktop app.
2. Run amp in a project directory, or use amp --no-tui on a remote machine.
3. Describe the task; the agent plans, edits files, runs commands and reports back.
4. Use the web or mobile view to steer long-running orbs, or open a portal to preview what was built.
5. For team work, move persistent threads into a space and use service accounts where the work should outlive an individual.

Pricing

Amp offers a free tier with limited daily usage plus paid credit plans for heavier agent use; team and enterprise options add shared spaces, service accounts and centralized billing. Check ampcode.com for current limits.

Links

• Website: ampcode.com
• Docs: ampcode.com/docs
• CLI remote control: ampcode.com/docs/cli/remote-control
• Changelog: ampcode.com/chronicle

Claude Code

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows -- all through natural language commands. Use it in your terminal, IDE, or tag @claude on Github.

Learn more in the official documentation.

Get started

[!NOTE]
Installation via npm is deprecated. Use one of the recommended methods below.

For more installation options, uninstall steps, and troubleshooting, see the setup documentation.

1. Install Claude Code:

   MacOS/Linux (Recommended):

   curl -fsSL https://claude.ai/install.sh | bash
   

   Copy

   Homebrew (MacOS/Linux):

   brew install --cask claude-code
   

   Copy

   Windows (Recommended):

   irm https://claude.ai/install.ps1 | iex
   

   Copy

   WinGet (Windows):

   winget install Anthropic.ClaudeCode
   

   Copy

   NPM (Deprecated):

   npm install -g @anthropic-ai/claude-code
   

   Copy

2. Navigate to your project directory and run claude.

Plugins

Anthropic's Claude Code plugins are developed in claude-plugins-official. See the plugins directory for more about plugins.

Reporting Bugs

We welcome your feedback. Use the /bug command to report issues directly within Claude Code, or file a GitHub issue.

Connect on Discord

Join the Claude Developers Discord to connect with other developers using Claude Code. Get help, share feedback, and discuss your projects with the community.

Data collection, usage, and retention

When you use Claude Code, we collect feedback, which includes usage data (such as code acceptance or rejections), associated conversation data, and user feedback submitted via the /bug command.

How we use your data

See our data usage policies.

Privacy safeguards

We have implemented several safeguards to protect your data, including limited retention periods for sensitive information, restricted access to user session data, and clear policies against using feedback for model training.

For full details, please review our Commercial Terms of Service and Privacy Policy.