A free ChatPDF that never uploads your documents An on-device ChatPDF, PDF.ai & AskYourPDF alternative.
Chatting with a PDF means trusting a server with the document. Feyard runs the entire retrieval pipeline — chunking, embedding, search and the answer — in your browser, so a contract, thesis or medical file stays on the device while you still get cited, page-grounded answers.
Comparison reflects publicly documented free-tier behaviour as of 22 October 2026. Confirm specifics against each vendor's own page before making a decision.
Your documents never leave the device
Indexing and question-answering run on WebAssembly / WebGPU locally. There is no upload endpoint, so confidential PDFs are never transmitted or retained by a third party.
Answers cite the exact pages
Every answer links back to the source passage and page number, so you can verify a claim instead of trusting a black box.
Chat across many documents
Drop in a whole folder of PDFs and ask one question across all of them — the engine retrieves from the combined corpus, free up to 5 documents (20 on Pro).
Follow-up questions that actually remember
Short follow-ups like “why?” or “what about the second one?” are rewritten into a full search query from the previous turn, and the last two turns are passed to the local model. Chats and their citations are saved in your browser, so reloading the page does not wipe your work.
Instant summary and key points, no model download
A one-click document brief is built by scoring and extracting the document's own sentences. It appears immediately, adds no hallucinated claims, and needs neither a 483 MB language model nor a server round-trip.
No account, no server retention
Open the page and start asking — there is no sign-up. Local files are never uploaded. Chat history lives in your browser's IndexedDB, not on our servers; clearing site data removes it. Only a public URL you explicitly paste is fetched through a hardened proxy, and its text is never logged or stored.
Feyard vs ChatPDF vs PDF.ai vs AskYourPDF
Every competitor below sends your files to a server and charges for the round trip. The differences that matter are where the data is processed and what the free tier is allowed to be used for.
| Capability | Feyard | ChatPDF | PDF.ai | AskYourPDF |
|---|---|---|---|---|
| Where documents are processed | In your browser — never uploaded | Cloud (upload required) | Cloud (upload required) | Cloud (upload required) |
| Upload required | No | Yes | Yes | Yes |
| Answers cite page numbers | Yes — source passages + page | Yes | Yes | Limited |
| Chat across multiple docs | Yes (5 free / 20 Pro) | Yes (paid tiers) | Yes (paid tiers) | Yes (paid tiers) |
| Free document allowance | 5 documents | Limited free | Limited free | Limited free |
| Account required | No | Yes | Yes | Yes |
| Multi-turn follow-up questions | Yes — query rewriting + last 2 turns | Yes | Yes | Limited |
| Chat history survives a reload | Yes — saved locally, no account | Yes (account required) | Yes (account required) | Yes (account required) |
| Instant summary & key points | Yes — extractive, zero model download | Yes (cloud model) | Yes (cloud model) | Yes (cloud model) |
| Import a web page by URL | Yes — hardened proxy, nothing stored | Limited | Yes | Yes |
| Scanned PDF OCR | Yes — on-device, 6 languages | Limited | Limited | Limited |
| Retrieval pipeline | Open: BM25 + vector RRF, multi-query, MMR | Undisclosed cloud pipeline | Undisclosed cloud pipeline | Undisclosed cloud pipeline |
| Source shown in original context | Yes — matched chunk highlighted among neighbours | Passage only | Passage only | Passage only |
| Price | $0 | Subscription | Subscription | Subscription |
RAG Search alternatives FAQ
Looking for something specific? The dedicated product page covers the engine, limits and batch export in detail — open the rag search.