Brethof AI Reviews

Claude Code review

One app for everything, but closed and paid

Not sponsoredAI Agents · Developer Tools

Tested on Linux · Paid

Claude Code is Anthropic's AI coding agent. In the desktop app's Code tab it reads your project, edits files, runs commands and tests, and shows the result. Chat, a terminal, a browser preview and parallel sessions share one window. We tested it on Linux, where it is in beta.

The good

  • Everything in one app. Claude Code puts chat, a terminal, a browser and teamwork in one app. With the previews, almost all the work happens in one place.
  • Many agents at once. The best part is the multi-agents: several sessions, and even agent teams, work in parallel.
  • Previews built in. The previews are one of the best parts: the app runs your project in its built-in browser, so you see the result without leaving it.
  • Routines and memory. It has routines, which are scheduled runs in the cloud, and built-in memory that it keeps by itself next to CLAUDE.md.
  • Other models too. It runs other models too, not only Claude. Ollama even has a one-line start for it, and can connect the desktop app as well, on macOS only. We didn't try either.

The not so good

  • Closed source. The real minus: Claude Code is closed source, so you don't know what's inside. Its licence reads all rights reserved.
  • Needs a paid plan. Claude Code needs a paid plan: Pro is 20 dollars a month, and heavy daily use means Max from 100 dollars a month, with 5-hour and weekly limits.

Our verdict

Recommended

Almost all the work happens in one app: several agents at once, a live preview of what they build, a terminal and routines. But it's closed source and needs a paid plan, so you trust Anthropic with what's inside and pay every month.

Reviewed: Claude Code 2.1.295 (desktop app 2.37337.1) · facts checked 2026-10-15

← All reviews

Hear it when it ships

New releases, real benchmarks and the occasional deep-dive. No spam, unsubscribe in one click.

Everything we build

External:   YouTube · GitHub