Option A: Claude: lists per-touch claims, you check, it reports
Best when: You want a correction gate between two AI passes
What it does: Claude lists and reports; you check and resolve
Strong at: Drafts, edits and summarises in many formats
Limit: You inspect both the working list and the report
Free plan: Yes, limited (5-hour and weekly usage caps)
Option B: ChatGPT: drafts the report from your hand-built list
Best when: You author the working list; the tool drafts the report
What it does: ChatGPT drafts the report; you check and resolve
Strong at: Compares and contrasts documents
Limit: An omission in your list still needs catching at the end
Free plan: Yes, limited (file uploads, images, voice, data analysis)
Option C: Gemini Notebook: lists per-touch claims, you resolve locally
Best when: You want AI help with the list, final report stays local
What it does: Gemini Notebook lists; you build and resolve the report
Strong at: Grounded answers with in-line citations
Limit: You write the final report after correcting the list
Free plan: Yes, limited (sources capped at 500,000 words or 200 MB)