Skip to content
MechPeptides

COA checks & dose maths

Check a report

Not a seller, not a lab Free, no account

Check a peptide lab report before you trust the vial.

A certificate proves nothing until it resolves against the lab that issued it. Free tools for checking a report, seeing who has published one, and doing the reconstitution maths.

Janoshik format check

Runs in your browser

Go straight to the lab's copy

Two values sit on every genuine Janoshik report. Enter them once and we hand them to the lab's own form; the answer is theirs, not ours.

Verify on janoshik.com
Answer comes from janoshik.com/verification/ [Nothing is sent to us]
201

Tests accumulated

80

Names listing

25

Most by one lister

30 AUG 2026

Last captured

Verification

How to verify a peptide COA

Every genuine Janoshik report carries a task number and a unique key. Enter both at janoshik.com/verification/ and the lab returns the report it issued. If the compound, quantity or date does not match what you were sold, the certificate belongs to a different sample. A report with either value cropped out cannot be checked at all.

  1. Find both values on the report

    A task number of five or six digits, and a unique key of twelve letters and digits. If a seller sends a certificate with either one cropped out, that is your answer.

  2. Send them to the lab

    Only janoshik.com/verification/ can say whether the report exists. Our field carries both values over to it, so you type them once rather than twice.

  3. Read what came back, carefully

    Match the compound, the quantity and the date against what you were sold. A real report for a different product is the most common trick there is.

Report anatomy

Format, not a specimen
Task number 216154 Five or six digits. Identifies the job, and on its own proves nothing: anyone can read a task number off someone else's report.
Unique key N4WSSXRV7EAN Twelve letters and digits. This is the half that proves the copy in front of you is the one the lab issued.
Those two fields are the whole game. Everything else printed on a certificate, including the lab's logo and the purity figure, can be retyped in a graphics editor.

The ledger

Who has actually paid for testing

Janoshik displays a rolling archive of the reports it has issued, along with the name whoever paid for the test asked to have printed on it. We track what it shows and group it by that name. It is the closest thing to a public record of who is willing to pay for testing, and it is not a recommendation.

49%

of the 80 names captured from the archive have published exactly one report. A single test on a single batch is the most common thing a lister has ever done.

  1. 01longevity pro's25
  2. 02INNOPEPTIDE12
  3. 03info@peptidegurus.com7
  4. 04ma@qianyecao.cn7
  5. 05vesuhealtheva@gmail.com7
  6. 06Yiwu Aozuo Trading Firm6
  7. 07MEI-PEPTIDE5
  8. 08HPEBIO4
  9. 09rachel.uwa.usa@gmail.com4
  10. 10RETA-PEPTIDE4
  11. 11SH Petitde Ltd4
  12. 12Anonymous3

The other 68 names account for 113 reports between them, and are listed in full on the ledger page.

201 tests from 80 listers, accumulated from public.janoshik.com's rolling window across 2 captures, 2026-08-29 to 2026-08-30. Spellings that differ only in case or spacing are counted as one name; anything else is left exactly as it was submitted.

Illustration: an unlabelled glass vial of lyophilised powder beside a U-100 insulin syringe.
A report describes the vial that was tested. Everything after that, including the one in your hand, is an assumption you are making.

Limits

What a verified report does and does not prove

Verification is a strong signal about one sample on one day. It is routinely sold as more than that, so it is worth being precise about the edges.

It does establish

  • The lab issued this report, and the numbers on your copy have not been altered.
  • The sample the lab received contained what the report says, at the purity stated.
  • Someone paid for testing, which is a cost most sellers of nothing will not carry.

It does not establish

  • That the vial you were shipped came from the batch that was tested.
  • That the seller listing the report is the party who submitted the sample.
  • That the product is sterile, correctly dosed, or safe to put in a human.

Pricing

What testing costs

Per sample, before shipping. Useful for judging whether a seller's report library was expensive to build or nearly free.

Janoshik published prices, checked 28 August 2026.
TestPrice
BPC-157 analysis$215
BPC-157 / TB-500 blend analysis$320
Common GLP-1 blind test (semaglutide, tirzepatide, retatrutide)$360
Typical single-peptide analysis (GHK-Cu, MOTS-c, cagrilintide, most others)$450
Human growth hormone panel (amount, purity, dimer)$500
CagriSema blend analysis$630

Compare methods and turnaround on the testing labs page.

FAQ

Common questions

Can I verify a report without the unique key?+

No. The task number alone tells you nothing, and a seller who supplies only the number has given you a report you cannot check. Some reports are findable in the public archive by task number, which is worth trying, but the key is what proves the copy in front of you is unmodified.

Does being in your ledger mean a seller is trustworthy?+

It does not, and the ledger says so on the page. It mirrors who has published a report in the public archive, with names copied exactly as they were submitted. Anyone can pay for one test on one good batch. Treat the ledger as a starting list of names to check, never as a recommendation.

Why does the calculator want the dose in milligrams?+

Because units are a property of your syringe, not of the drug. The same 2mg is a different number of units on a 0.3mL and a 1mL U-100 syringe, so a dose described in units is only meaningful to whoever owns that exact syringe. Milligrams in, units out, arithmetic shown.

Do you sell peptides or take commission from sellers?+

We do not sell peptides, and no link on this site earns a commission today. If one ever does, the link itself will be marked as paid where it appears, rather than mentioned once in a policy nobody reads.