I switched from ChatGPT to Claude for my writing work and the difference in how it handles long documents surprised me

I write long form content for a living, research reports, white papers, detailed guides. I had been using ChatGPT as my main AI assistant for about eight months. I switched to Claude three months ago mostly out of curiosity and stayed because of one specific thing.

When I paste in a long document and ask Claude to work with it, it actually seems to hold the whole thing in mind. I can refer back to something from page one while we are working on something from page eight and it tracks correctly. With ChatGPT I was constantly getting responses that clearly did not account for something I had given it earlier in the same conversation.

I want to know if this is a real architectural difference or if I have just been prompting differently without realising it. And for people who do heavy document work, what has your experience been comparing the two?