Claude vs ChatGPT vs Gemini vs Copilot for writing

No assistant wins everywhere. Pick by task, and remember these capabilities move fast: the table is a snapshot, not a verdict. All four are competent writers. The differences are about fit.

The honest summary: the deciding factor is usually where your documents live, not which model drafts marginally better. If your work is in Word, Outlook and Teams, Copilot’s in-context access beats a slightly better draft you have to copy-paste. If it is in Docs and Gmail, Gemini has the same advantage. If you care most about tone and following detailed instructions, Claude holds a brief tightly. ChatGPT is the flexible generalist with the widest ecosystem of extras.

How to read the table

“Strong” and “Good” describe typical results on everyday writing, not benchmark scores. A dash means the tool doesn’t do that thing directly. Claude and ChatGPT don’t live inside your files the way Copilot and Gemini do, though you can paste content in.

“Native” is doing real work in that row. It means the assistant can already see the document you’re editing and the mail in your inbox, without you moving anything. That convenience is the single biggest practical difference on this table.

The comparison is getting blurrier

The assistant and the model underneath it are no longer the same choice. Microsoft 365 Copilot now offers both OpenAI and Anthropic frontier models in some applications, so “using Copilot” no longer tells you which model answered. Expect more of this. The useful question is shifting from which assistant to which model, in which place, with access to what.

Don’t over-think the choice

For most writing, the prompting habits matter more than which of the four you use. Reach for the one already in your workflow, and switch only when a specific task clearly favours another: tight rules, files-native editing, or heavy multimodal input.