The State of Peptide Testing in 2026: What 9,997 Published Results Show

When this report was compiled in September 2026, PeptideCensus indexed 9,997 published third-party test results spanning 97 compounds and more than 400 ranked vendor-compound records. The index refreshes daily from the public testing record, so today’s totals are higher; every figure below is that September snapshot. This is what that dataset says about the research peptide market in September 2026. Every number below comes from published, traceable results; the methodology and its limits are at the end.

The most-tested peptide has never been approved

Retatrutide accounts for 3,694 published tests, 37% of the entire record, more than tirzepatide (2,401) and semaglutide (342) combined. An investigational compound that no one can legally prescribe is the most heavily quality-tested substance in the gray market. The mechanism is simple: where there is no approved product, the entire demand flows through research vendors, and testing follows demand.

Semaglutide inverts the logic. It is the most famous molecule of the decade and one of the least-tested majors in this record, because the approved product absorbs most of its demand.

Two-thirds of all testing is one drug class

GLP-1-family compounds (retatrutide, tirzepatide, semaglutide, cagrilintide, survodutide, mazdutide) account for 6,634 of 9,997 tests: 66.4%. The other 91 compounds we track share the remaining third. Practical consequence for buyers: outside the GLP-1 class, per-vendor testing records get thin fast, and a thin record deserves less trust however good its numbers look.

Where the testing concentrates

A third of ranked records are stale

Of 162 dated vendor-compound records in our leaderboards, 54 rest on a most-recent test older than six months, and 25 are older than a full year. Staleness is not evenly distributed: tirzepatide leaders retest within weeks, while the entire top of the melanotan II table runs on mid-2025 results.

A score is a claim about batches that existed when the tests ran. The older the test, the less it says about the vial that ships tomorrow, which is why every leaderboard on this site carries a last-tested column next to the score.

Price says almost nothing about quality

The median published price across all ranked records is $1.00 per milligram. Around that median, the spreads are enormous and mostly uncorrelated with test results. Semaglutide runs a 28-fold spread between its cheapest and most expensive scored vendors. Ipamorelin’s top five vendors span $0.72 to $3.74 per milligram while their scores sit within three points of each other.

And the compound with the best scores in our entire index, GHK-Cu, is also nearly the cheapest at $0.05 to $0.10 per milligram, because a tripeptide is trivial to manufacture well. What price does reliably signal is positioning: bulk-oriented sellers near raw-material cost, retail storefronts several multiples higher, same molecule.

High scores are the exception, and that is the point

Only 44 of 163 scored records, 27%, clear a score of 80. Sixty sit below 60. Some of that reflects genuine label failures, which turn up regularly in Janoshik’s public results database. But much of it reflects evidence depth, because our scoring is confidence-adjusted: a vendor with four clean tests cannot score like a vendor with twenty-six.

Entire compounds illustrate the cap. No CJC-1295 vendor clears 48, not because every batch fails, but because nobody has published a record deep enough to earn trust. Read low scores as unearned trust at least as often as bad product.

How is this data built?

We aggregate publicly published third-party test results, principally records traceable to Janoshik Analytical’s open verification system and the public dataset Finnrick maintains, and recompute vendor scores daily. We test nothing ourselves and sell no placement in any ranking.

One structural caveat applies to every number above: publication is opt-in, and nobody publishes failures they can quietly withdraw, so the public record skews positive by construction. Aggregation across many batches and vendors is the corrective, and it is the entire reason this site exists. Sources and verification tools are listed on the data sources page; any individual report can be checked with our verification guide.

Common questions

How many peptide lab tests are publicly available?

In the September 2026 snapshot behind this report, our index tracked 9,997 published third-party results across 97 compounds; the live count grows daily. The true number of tests performed is higher; unpublished and withdrawn results never enter the public record.

Which peptide has the most published testing?

Retatrutide, with 3,694 published tests across 256 tracked vendors as of September 2026 (4,567 today), more than tirzepatide and semaglutide combined.

What share of peptides fail testing?

No verified market-wide figure exists. The often-quoted “43% of Janoshik’s 2024 submissions” has no primary source we could find, so we don’t use it. Failed results do turn up regularly in Janoshik’s public results database, and even that undercounts, because vendors choose which results get published.

Does a higher price mean a better peptide?

The data says no. Within-compound price spreads reach 28x with little score correlation, and the best-scoring compound in our index is among the cheapest.

Use the record, not the marketing

Every claim above is traceable to a published test, which is the standard we hold vendors to as well. Start from the vendor rankings, open the compound you research on the peptide index, and check the batch before the cart.