How to Compare Three Vendor Quotes With ChatPDF
ChatPDF lets you ask a question across several documents in one session, which sounds like exactly what a three-quote comparison needs. Half true. It reads each PDF correctly and cites properly, but it will not normalize differently-structured quotes into a single table. You ask for the same field from each file, then build the grid yourself.
That small correction to expectations turns a frustrating afternoon into a fifty-minute job. Here is the sequence.
Can ChatPDF read multiple PDFs at once?
Yes on paid plans, where a session can hold several documents and a question can span them. The free tier is more restrictive: fewer documents per day, fewer questions, and no cross-document collection. If you need three quotes side by side, plan on a paid month or run the comparison one file at a time and collate in a spreadsheet.
Per-document answers are just as good on the free tier. What you lose is the ability to ask a single question across all three.
What fields should you extract from every quote?
Fix the schema before reading anything:
- Unit price and currency
- Volume or user count the price covers
- Payment terms in days
- Contract length and renewal behavior
- Setup or onboarding fee
- Support tier and response time
- Exclusions buried in terms
The exclusions line is where quotes differ most and where buyers lose money. Ask each file directly: “What is explicitly excluded from this quote?”
Why does it refuse to build a comparison table?
Because each supplier structures a quote differently. One lists per-seat pricing, another bundles a platform fee plus usage. The model retrieves text blocks; it does not reshape them into a canonical schema across files. When you ask for a combined table you get something that looks right and silently compares per-seat to per-usage rates.
Ask per-document, then assemble. It is slower and correct.
The comparison grid is your work. The tool’s job is fetching the right number off the right page.
How do you catch the clauses that change the ranking?
After prices, ask each vendor document three questions: “What happens if we exceed the stated volume?”, “When can the price change?”, and “What is the notice period to cancel?” A quote that looks 15% cheaper with a 60-day price-change clause and no volume buffer is often the more expensive option at renewal.
Put those answers in a column next to price. Decision quality jumps.
| Step | Time | Who does it |
|---|---|---|
| Extract per-document fields | 10 min | ChatPDF |
| Verify two numbers per file | 5 min | You |
| Build and label the grid | 20 min | You |
| Ask about exclusions and escalation | 10 min | ChatPDF |
| Note the decision and rationale | 5 min | You |
What are the limits for procurement work?
Scanned quotes with a signature image and no text layer need OCR first. Anything in an attached Excel sheet is invisible. And if two quotes are from the same vendor with different dates, label the files clearly, because the answer cites the page, not the filename you remember.
For contract-level clause comparison specifically, ChatDOC has a two-document compare panel that flags changed paragraphs, which is a better fit for redlines than asking ChatPDF questions.
Quick Summary
- ChatPDF reads several PDFs per session on paid plans, with per-page citations.
- It extracts fields well per document but will not normalize quotes into one table.
- Fixed extraction schema plus a self-built grid beats one prompt asking for a comparison.
- Exclusions, volume overages and price-change clauses decide more deals than unit price.
- Not suitable for attached spreadsheets or scanned quotes without OCR.
