Skip to content

PepAssure About

An independent rating of research-peptide vendors

PepAssure scores peptide vendors from lab results published by third parties, using one formula that we publish in full. We sell no peptides, and no vendor can pay to change its score.

lab results indexed
3,601
Third-party results matched to a vendor, each able to move its score.
vendors with lab results
543
At least one of those results is attached to each.
vendors scored
446
242 more are listed as “Not yet scored”.
compounds covered
149
Distinct peptides and blends across those results, doses merged.
Newest test date on file
15 Nov 2026
Last scoring run
15 Nov 2026, 05:36 UTC

The problem

Most research peptides are sold as research chemicals, outside the approval and inspection that apply to medicines. A buyer cannot see what is in a vial before it arrives.

The usual proof is a certificate of analysis that the vendor chooses to publish. Independent results do exist, but they are scattered across sites and hard to compare. The difference between the two comes down to one question: who chose the sample?

A certificate the vendor publishes
  1. 1Vendor picks the batch
  2. 2Vendor picks the lab
  3. 3Vendor decides what to publish

The vendor chooses every step.

A rating built from bought samples
  1. 1Buyers or the rater buy from the store
  2. 2A lab tests what arrived
  3. 3The rater publishes the results

The vendor does not choose the sample.

Finnrick, the source behind most of our results, says its samples come from public submissions and its own purchases, that vendor-submitted samples are labelled as such, and that vendor-supplied CoAs do not drive its rankings.

What we do

  • We index third-party results. Each night we re-read a rotating batch of Finnrick's public ratings and match each result to the vendor it names. Janoshik certificates and reports that vendors upload from recognized labs are added too, at the full or half weight the methodology sets out.
  • We score every vendor the same published way. One formula, with its weights, caps and penalties written out on the methodology page, alongside a calculation reproduced from live data.
  • We withhold a score rather than guess. A catalog vendor with no counted lab result is shown as “Not yet scored”. The exceptions, provisional scores for vendors who claim a listing or that a lab index or the community brought to us, and scores our team has pinned by hand, are set out and counted in the methodology.

How we are paid

PepAssure is paid by vendors, for subscriptions and add-ons: referral analytics, price listings, competitor benchmarking, replies to reviews, API access, a printable verification report, and help uploading lab reports. Claiming a listing is free. Plans and prices are on the For Vendors page.

When we get it wrong

Vendors can request a rescore from their dashboard. Anyone can email hello@pepassure.com with the vendor, what is wrong and a link to the evidence. A score changes when the evidence does, never because of a complaint or a payment. The full process is under corrections and disputes.

Who we are

PepAssure is run by a small independent team. We do not publish individual names; judge the work by its evidence and its method, both of which are on this site.

Method record

Dated changes to how we score, as recorded in our code history. We list nothing we cannot date.

  1. 14 Nov 2026

    Certificates from Vanguard Laboratory and MZ Biolabs now count, at half the weight of a blind test, like every vendor-supplied certificate. A pillar with no data now withholds 5% of its weight instead of 15%. When a vendor has a large library of recognized-lab certificates next to a few blind tests, the certificates may carry up to 70% of CoA Verification (half up to 10 certificates, then a tenth more each time the library doubles); the blind tests always keep at least 30%.

  2. 26 Oct 2026

    X posts and Trustpilot stars can move Community Sentiment when a vendor already has a lab result or a provisional score. A Trustpilot review that reads like an ad still counts by its stars. Reddit mentions stay on the vendor page and still do not score.

  3. 20 Oct 2026

    Certificates catalogued by PeptideBenchmark now count toward Purity at half of a Finnrick result, unless the testing lab named on the certificate is Finnrick or Janoshik, in which case the result keeps full weight. They are not a blind test, they do not enter CoA Verification, and a vendor name counts only when it matches our catalog at 0.7 or better.

  4. 12 Oct 2026

    Freedom Diagnostics certificates are now read directly from the lab's own public index, with the client name on the PDF checked against the vendor. They count at the vendor-submitted weight. Where Peptigrity catalogued the same certificate, the lab's copy counts instead; a Peptigrity certificate the lab's index does not know is set aside.

  5. 12 Oct 2026

    A lab is now recognized by one rule: it runs a public certificate lookup anyone holding a report can use. Five labs that pass were added (ILS, Bioviridian, Ethos Analytics, TrustPointe, MDx BioAnalytical), so their certificates count at the vendor-submitted weight; every vendor page links the lab's lookup next to such rows.

  6. 10 Oct 2026

    The methodology page was rewritten to describe only what the scoring code does: each pillar now lists what feeds it and what we do not do.

  7. 10 Oct 2026

    Scores for catalog vendors are now published every night from the evidence on file. A vendor with no qualifying evidence is withheld as “Not yet scored” instead of keeping an old number.

  8. 10 Oct 2026

    Bulk powder manufacturers and trading companies are excluded from the rankings. A supplier is classed as wholesale only when at least two signals on its Finnrick page point that way, such as bulk per-milligram pricing or a company-style name with no retail storefront.

  9. 10 Oct 2026

    Pillar values that 21 catalog vendors carried from launch, with no measurement behind them, were removed. Those vendors are unscored until lab evidence exists.

  10. 10 Oct 2026

    Reports a vendor uploads from a recognized lab now count toward Purity and CoA at half weight, once a person has read the purity off the document.

  11. 9 Oct 2026

    Claimed vendors get a score coach in their dashboard that explains which inputs would move their score. It gives advice only; the score is computed the same way for every vendor.

  12. 9 Oct 2026

    The score breakdown on vendor pages shows the five real pillar values and the date they were computed, instead of an empty state.

  13. 27 Aug 2026

    A discovered or nominated vendor with no evidence is left unscored instead of receiving a default score of 80.

  14. 27 Aug 2026

    Each lab result is credited to the lab named on it. Certificates a vendor published itself had been captioned as Janoshik results.

  15. 11 May 2026

    The CoA pillar reads the lab-results index: Finnrick and Janoshik blended 70/30, weighted by test age, with Janoshik-only results capped at 85.

  16. 11 May 2026

    Finnrick ratings joined the lab-results index that scoring reads.

Contact

Questions, corrections and disputes
hello@pepassure.com
Help with a vendor account or billing
support@pepassure.com

Read the methodologyBrowse scored vendorsContact form