
Best thing about Claude is Claude Code -its integration with the bare CLI is amazing. If you love working in the terminal, it’s great to use. The app and web UI are good too, but I personally prefer using it in the terminal.
In the terminal, its performance is great, and the context window is solid as well - especially since it was recently increased with their new partnership. New AI models like Opus 5.5 perform much better and use far fewer tokens than older ones like 4.8. It also now supports Agents.md along with Claude.md, which makes it easier to manage context across multiple LLMs in a shared project. On top of that, it supports various plugins, so you can onboard and get started quickly without writing custom skills or anything like that. Review collected by and hosted on G2.com.
There’s not much to dislike, but there are a few nice-to-have features that still feel missing in Claude - like image generation and an in-app preview window similar to Codex/ChatGPT, where you can preview code changes, MDs, and the browser. Some of this exists in the Claude app, but it isn’t as intuitive as it is in the Codex app. Review collected by and hosted on G2.com.