Nolen began their writing career in 2019, with three years dedicated to editing the Creative section at MakeUseOf. Their expertise lies at the crossroads of technology and creativity, covering areas like photography, video editing, and graphic design.
Outside of work, you'll often find Nolen diving into a good book, writing their own stories, or playing video games.
Most AI-generated files end up living somewhere completely different from where they were created. A chatbot can draft your task list or spec sheet, then you put it in a Google Doc, owned by a completely different company. Wherever you put the notes, that doc ends up with no idea about the conversation behind it. Then Anthropic launched its own editor inside Claude in September, which completely cuts out the middleman between AI synthesis and note-taking. I've been using it in regular chats and in Claude Code, and I think it might have started replacing Google Docs for me.
What is Claude Docs?
It's not Google Docs wearing a Claude badge
Claude Docs is Anthropic's own rich-text editor, which launched alongside Claude Slides in September, and it isn't a Google product in any form. It actually confused me a bit at first because I'm so used to associating the word "Doc" with Google Docs, and there are also multiple things sharing the same name in Claude right now: Anthropic's developer documentation, the connector that edits your actual Google Docs, and Claude Docs.
A doc lives in your Claude account, and whatever you make will end up in the Artifacts tab, so you can get back to that doc from any conversation. Claude Docs is available on all the tiers, including Free, and it's also switched on by default. An important distinction is that Claude Docs is separate from the file creation that Claude has had for ages. For example, a .docx that Claude generates is finished once you download it. But a Claude Doc stays open and editable by both you and Claude.
There are a couple of surfaces where you'll encounter docs in Claude. Firstly, you've probably noticed lately that responses in the chatbot end with a "Create Doc" option at the bottom. You can also ask the chat at any point to make a doc, or just type the /doc command if you don't want Claude to guess what you want. When opened to full view, the doc gets its own menu where you can pop it out into a separate window or open it in your browser, plus you can split one into tabs like sections of a notebook. And as I've mentioned, they also live in Artifacts - so you can go straight to the Artifacts tab and create one from scratch.
Everything I used in Google Docs is also in Claude Docs
It covers pretty much everything, and more
The first time I asked for a doc, Claude didn't just hand me a finished page. It put up an outline and then filled in the sections one at a time while I could watch it live, and since it's not locked while it's at work, I could also type into the doc alongside it. Hitting the slash brings up the block menu with all your usuals like headers and list options, and you've got a couple more formatting tools in the top toolbar such as tables and image inserts.
The comments are what make it worth even creating a doc. You highlight anything in the doc and tag @claude, add your commentary or ask Claude a question or for revisions. It will respond in that thread once it's made the edit, with a short note on what's changed. There are quick buttons for making sections shorter or simpler as well. The thing is, Claude responds to all of these little interactions with the full context of the session in mind. So you're not just slapping AI on top of the notes, you're interacting with the entire chat in an organized fashion, basically.
At first I didn't realize it had version history, but, similar to Google Docs, it's tucked away in the top-left dropdown; this will show you a list of all the edits. Then there's also export. You can get out your document as a Word file, PDF, Markdown file, or send it straight to Google Docs or Notion. This is ultimately what makes Claude Docs a true Google Docs replacement because you can slot this into pretty much any workflow with any tool stack.
The one area it trails behind is sharing. Whoever opens your link needs a Claude account, and there's no comment-only role for people who should only leave notes. I'd also be careful with deleting, since there's no trash to pull a doc back out of.
It already knows your projects before you type anything
All it needs is a sentence and a folder
Docs in regular chat can already draw on your files and memory, plus whatever apps you've connected, but Claude Code is where this gets super useful. A Claude Code session already sits inside a folder, so when you ask for a doc it reads the actual files instead of working off whatever you've said in the chat. Plus it knows everything that's happened in that session.
Here, I opened a session in a copy of my Obsidian vault and typed one sentence asking for a status brief, with a table of what's active and what's stalled, without pasting in any notes. It built an empty skeleton first, with placeholders where each section would go, and asked permission before editing the doc. Then it read through my files and filled everything in.
My vault has dashboards that are supposed to track all of this, and Claude noticed they hadn't been regenerated in a hot minute, so it just ignored them and sorted things by each file's real last-edited date instead. The status column used little colored labels you can click to change, like Active or Stalled, instead of plain text. Further down, the next steps were a checklist I could actually tick off.
A blank Google Doc would've left me digging through the vault myself to fill in that table. Here the doc already knew the whole project before I typed anything.
Starting from context instead of nothing
Google Docs isn't disappearing for me entirely, mainly because other people still send me links to projects and briefs. It also has Gemini which can handle comments and it can dig through my Drive, but turning a proper Gemini conversation into a doc still means building it in the Gemini app and exporting it over. This is why I keep starting in Claude Docs now instead, everything is properly integrated.
