Four groups now publish testing data on research-peptide products sold online, and they don't all measure the same thing. Finnrick scores individual samples and rolls them into A-E vendor grades. VialAudit buys anonymously and prints the chromatogram. Peptigrity aggregates lab tests against community reviews. Underneath most of these sits an analytical lab, frequently Janoshik Analytical in the Czech Republic, running the actual HPLC and mass-spec work. Reading the four together is more useful than trusting any one, because their methods have different blind spots.
This is a report of what these raters have published. It is not an endorsement of any product or vendor, and it is not medical or dosing advice. Most compounds discussed are sold for laboratory research use only and are not approved for human consumption.
What each rater measures, and how
The scoring schemes are public, which is the first thing worth checking.
Finnrick's methodology splits a 0-10 sample score into three parts: purity (0-4 points), quantity accuracy versus the label or batch claim (0-4 points), and batch-information quality (0-2 points). A sample at 99.9% purity earns the full 4; below 98% earns 0-1. On quantity, within plus or minus 5% of the claimed amount earns 4 points, while more than 20% off earns 0-1. Letter grades require clearing both an average and a minimum across at least two tests: an A needs an average of 7 or more and a minimum of 6, an E is the floor, and F is reserved for documented, widespread fraud. As of mid-2026 the site reports 8,026 samples from 225 vendors across 15 peptides.
VialAudit runs a different protocol: anonymous orders under an alias, then samples rotated between two labs it deliberately doesn't name "so vendors can't pre-screen samples by lab signature," with quarterly cycles on every vendor it covers. It describes itself as "a publication, not a vendor" with zero sponsored placements. The point of buying blind is to defeat the main weakness of vendor-supplied certificates.
Peptigrity weights its score 50% lab purity and 50% community reviews, and as of this writing lists 7,906 lab tests across 64 peptides and 345 shops. It leans on the volume of published per-test HPLC results rather than its own buying.
Purity: high on paper, with a selection-bias asterisk
The headline pattern across all of them is that published purity for the major peptides clusters high. Recent Peptigrity entries show figures like 99.8% NAD+, 99.9% ipamorelin and 100.0% retatrutide for specific batches. When VialAudit audited a 200-test mirror of Janoshik's public database, the median purity came out at 99.7%.
That number is the trap. VialAudit is explicit that 99.7% "reflects selection bias rather than industry-wide performance." The public certificate databases are vendor-curated: a seller chooses which batch to test and which result to post. Of the 200 tests VialAudit examined, 195 passed identity confirmation, five carried identity flags, and one had an explicit pass=false marker, so failures do appear, but the dataset is still the vendors' own highlight reel. The audit's blunt conclusion is that cherry-picking "isn't caught by external verification alone" and that independent retesting is the only real check.
VialAudit's check on the lab itself is the more reassuring finding. It verified that Janoshik's certificate metadata matched the report contents across all 200 tests, that QR codes resolve to Janoshik's own servers (so a fabricated PDF won't verify), and that negative results sit publicly alongside passing ones. The instrument data appears to be real. What it covers is the narrow question of a single submitted vial.
Identity and the underdosing problem
Identity failures are rarer than dose problems in this record, but they exist, which is why both Finnrick and the labs run mass spectrometry alongside HPLC. The bigger and more systematic issue is quantity. Finnrick treats it as fully half the technical score for a reason: a vial can be 99% pure and still contain far less active compound than the label says.
The mechanism is laid out in Peptigrity's purity-standards explainer. HPLC purity is "the percentage of UV-absorbing material in a sample that is the target peptide," measured by reverse-phase HPLC at 214 nm. It includes peptide-related impurities like deletion sequences and oxidation products. It excludes water, salts, TFA counter-ions, residual solvents and endotoxins. So purity is not the fraction of active compound in the vial.
The worked example is the part buyers miss:
- A vial labelled 5 mg at 99% HPLC purity but 75% net peptide content holds roughly 3.75 mg of actual peptide.
- That is a 25% concentration error against the label, despite a near-perfect purity figure.
- TFA counter-ions alone can be 10-25% of dry weight; residual moisture another 2-10%.
This is exactly the gap a purity-only certificate hides and a quantity assay catches. It also runs both directions. Underfill shortchanges the buyer; overage (more than labelled) shows up too, and Finnrick penalises any deviation beyond plus or minus 5% regardless of sign, because an unknown amount is as much a problem for controlled research as a low one.
| Rater | What it buys/tests | Purity | Quantity | Reviews | Stated scale |
|---|---|---|---|---|---|
| Finnrick | Public submissions + own investigations | 0-4 pts | 0-4 pts | no | 8,026 samples / 225 vendors |
| VialAudit | Blind anonymous orders, 2 rotating labs | yes | yes (label accuracy) | qualitative | quarterly cycles |
| Peptigrity | Aggregates published lab tests | 50% | within tests | 50% | 7,906 tests / 345 shops |
| Janoshik (lab) | Vendor-submitted vials | HPLC | mass/quantity assay | no | public verifiable DB |
What the aggregate actually supports
Read together, the raters support a few claims and undercut others. Published purity for popular peptides is genuinely high, but it is drawn from self-selected batches and should not be read as a market-wide pass rate. Identity failures are uncommon but real, and they only surface when someone runs mass spec rather than purity alone. Quantity is where the published data does the most work, because a clean purity figure can sit on top of a 25% dosing error.
The methodological split is the real signal. A vendor-curated certificate (the Janoshik public model) answers "is this one vial what they say," while a blind-buy audit (VialAudit) answers "is what they ship what they post." Those are different questions, and the gap between them is where most of the disagreement in this field lives. None of these numbers establishes safety or efficacy in humans, and none of the compounds is being described here as approved for that use.
Reporting of published third-party testing data, attributed to the named raters. Not medical, dosing, or purchasing advice.