Option A: ChatGPT: groups garments by compatible labels
Best when: You want groups from a row-by-row transcription
What it does: ChatGPT tables and groups; you check against labels
Strong at: Table built from your notes, nothing from memory
Limit: Accuracy still to confirm; check against originals
Free plan: Yes, limited (free tier limits images, voice, deep research, Codex)
Option B: Claude: resolves uncertain garments one at a time
Best when: You want uncertain garments resolved one at a time
What it does: Claude asks questions, drafts; you inspect and answer
Strong at: Conservative table with reasons linked to each label
Limit: Accuracy still to confirm; check against originals
Free plan: Yes, limited (usage capped in 5-hour and weekly windows)
Option C: Gemini: audits loads you already proposed
Best when: You want your own sorting plan audited for conflicts
What it does: Gemini compares and revises; you check each conflict
Strong at: Each change explained by its source text
Limit: Accuracy still to confirm; check against originals
Free plan: Yes, limited