How to read this page
Every fact about VerifyPDF below carries the date it was read and names the page it was read on: their pricing page and their home page, both read on 2026-08-06. Prices in this market move.
No detection rate is repeated here as evidence. Where their accuracy claim appears, it appears as a claim, in the same sentence as the fact that no methodology is published for it. We publish no detection rate of our own, for the same reason.
Side by side
| Tamperlens | VerifyPDF | Stronger | |
|---|---|---|---|
| Entry price | Free: 50 documents a month, with an API key, no card (/pricing, 2026-08-06). | Basic: $15 a month billed annually, $19 month-to-month, for 150 documents a month, single user. 15-day trial, no card (read 2026-08-06). | — |
| Under 1,000 documents a month, no API needed | Free covers 50. Above that the next step is Solo at $29 a month for 1,000 documents: more than you need, at more than you need to pay. | Basic at $15 a month covers 150 documents through their web tool. For that buyer this is simply cheaper than us (read 2026-08-06). | VerifyPDF |
| API access | On every plan, including Free. The API is the product; the web pages are a demonstration of it. | Not on Basic. REST API from Professional, $79 a month billed annually or $99 month-to-month, 1,000 documents (read 2026-08-06). | Tamperlens |
| Price per document at 1,000 a month | $29 for 1,000, which is $0.029 a document (2026-08-06). | $79 for 1,000 on annual billing, which is $0.079 a document; $0.099 month-to-month. Prepaid packages are offered for additional volume at discounted rates, but no per-document rate for them is published on the pricing page (read 2026-08-06). | Tamperlens |
| Known-good issuers | Issuer baselines built from genuine specimens, but no allowlist of banks: an unknown issuer is simply unknown to us. | A "Trusted" indicator for documents already verified in their system, "such as bank statements from major banks in the US and Europe" (read 2026-08-06). | VerifyPDF |
| Identity documents | None. We do not verify a person, and we do not read an identity document as an identity document. | Passports, driving licences, national IDs and residence permits, from a photo or scan in JPG, PNG or HEIC/HEIF (read 2026-08-06). | VerifyPDF |
| Speed | No published time. A 15-second hard deadline per document is the ceiling, after which the answer is an error rather than a wait. | "Checking a document typically takes 5 seconds, with longer multi-page documents taking up to 30 seconds" (read 2026-08-06). | VerifyPDF |
| What happens to the file | Held in memory, analysed, discarded with the response. Nothing about the document survives the request (/security). | Stored: "All documents are encrypted in transit and at rest", deleted "when the retention period expires", GDPR compliance stated. The retention period itself is not stated on the pages we read (read 2026-08-06). | Tamperlens |
| Accuracy claim | None. We publish no detection rate, because we have not measured recall. | "we are 99% accurate with computer-generated documents such as bank statements, payslips … tax returns … and invoices": no methodology, population or ground truth is published for this figure (read 2026-08-06). | — |
| False-positive rate | Published at /evidence, with its population and date. | None published (read 2026-08-06). | Tamperlens |
Volume decides which one is cheaper
This figure said "the gap between is theirs" until 2026-08-15, conceding the band their Basic plan sits in. It was wrong twice. A prepaid pack already covered that band at $0.05 a document against their $0.10, and it was not drawn here at all. Then Solo moved from $39 to $29 on the same day, which covers the band outright: 1,000 documents for $29 against 150 for $15, and the same engine either way. What stays true, and is the reason a concession was made at all, is that theirs is a subscription with a web tool for a single seat and ours is API-first: if you want a browser and a login rather than a key, neither the plan nor the pack answers you. Tamperlens plans from /pricing, 2026-08-06; the pack from the same page. VerifyPDF plans read 2026-08-06: Basic $15 a month billed annually (or $19 month-to-month) for 150 documents, single user, no API; Professional $79 billed annually (or $99 month-to-month) for 1,000 documents with REST API. Only published monthly plans are drawn, at the annual-billing price on both sides. Price is not capability: the rows above differ in what they include.
Where VerifyPDF is the better answer
- You check a few hundred documents a month by hand. At 150 documents with no API, $15 a month beats anything we sell. Our free tier stops at 50 and our next plan is built for a thousand.
- You need identity documents read. Passports and driving licences are a category we do not touch.
- You want a known-good list. An allowlist of issuers answers "is this the shape a real HSBC statement has" in a way that structural analysis alone does not. It is the capability gap in this comparison we would most like to close.
On images the two products are closer than they look: we read JPEG, PNG, WebP and HEIC as well, with nine image signal families. The difference is not the file type, it is that they identify documents and we do not.
Where Tamperlens is the better answer
- You are writing software. An API key on the free tier means you can integrate and test before anyone approves a purchase.
- Volume. At a thousand documents a month the per-document price is about half, and it keeps falling on the plans above.
- You cannot retain the document. We keep nothing after the response, which is a different regulatory conversation from "encrypted and deleted later".
- Two things nobody else in this band ships: recovering the words underneath a failed redaction by replaying paint order, and finding instructions hidden in a document to steer whatever model reads it next. See /redaction-check.
- Brazil. CPF, CNPJ and boleto identifiers are checked by arithmetic on the check digits, looking nothing up about anybody, and prices are quoted in reais.
What we cannot do
- No identity proofing and no identity documents.
- No issuer allowlist. See above; theirs is a real advantage.
- Recall is unmeasured. We know how often we flag genuine documents. We do not know how often we catch real tampering.
- A cleanly regenerated PDF launders most byte-level signals, and a document forged from scratch and never edited may carry no structural trace at all.
What none of us publishes
No vendor in the table above publishes a false-positive rate on legitimate documents: how often the system flags a genuine file. A 99% accuracy claim is not that number, and no methodology is published for it.
The four cells are the definition of a confusion matrix, not a measurement of either product. Their claim ("we are 99% accurate with computer-generated documents such as bank statements, payslips … tax returns … and invoices") is published with no methodology, population or ground truth (read 2026-08-06), so which cells it counts cannot be established from outside. Ours is at /evidence, and covers only the shaded cell.
Ours is at /evidence, with the population it was measured on, the date, and the engine version that produced it. The other comparisons are at /alternatives.
Read a report before you believe a comparison
Open the “Edited after creation” sample: a precomputed report on a document whose printed total was lowered by a second revision appended after creation. No account, no upload, no quota, and nothing on this page has to be taken on trust to look at it.
Tamperlens reports risk signals, not authenticity verdicts. Signals can have benign causes; combine them with your own decision logic.