PeptidesGrid doesn't test products. It has no lab, doesn't buy samples and doesn't grade vendors. The lab scores on the site come from Finnrick Analytics, a US testing service that has customer-submitted samples tested by third-party labs. Each score is shown with a link to the vendor's Finnrick profile.
What Finnrick publishes
When the site last fetched its data on Sep 25, 2026, Finnrick listed 1,884 vendors, 111 products and 10,405 tests. A vendor's public profile page, which needs no account, shows:
- an overall rating, and a rank among the vendors it ranks (345 at the last fetch);
- pass and fail counts, the number of tests and peptides tested, and the latest test date;
- a rating for each peptide tested at that vendor, with the number of tests;
- test history rows with the date, the peptide and pass or fail;
- the vendor's location and a status, such as Trading;
- notes, including warnings about fraud, sites using the vendor's name, or service problems;
- contact channels, some marked "Not verified".
Lab names, purity percentages, measured amounts, batch IDs and the original lab reports are for signed-in or paying members. PeptidesGrid uses only the public profile data and links back to it.
The "Purity focus" score
Finnrick sorts each test into pass or fail using a set of rules it calls a classifier. The score on PeptidesGrid uses Finnrick's public default classifier, "Purity focus". Under it, a test passes when both of these hold:
- purity is 99.5% or better;
- the measured amount of peptide is within 35% of the amount on the label.
The score is the share of classifiable tests that pass, adjusted so that a vendor with few tests doesn't look more certain than it is. It is not a raw pass rate. Tests that were incomplete or couldn't be classified count toward a vendor's test total but not toward its score. A vendor with no classifiable tests has no score, which says nothing either way.
Finnrick has other classifiers that weigh purity, the measured amount, batch information and, in one case, endotoxins differently. Scores from different classifiers aren't comparable, so the site shows Purity focus only.
Why few tests pull a score toward the middle
With few tests, chance matters a lot. A vendor whose only sample passed has a 100% raw pass rate, and one failed sample would halve it. Finnrick's adjustment blends each vendor's results with a typical value and gives the vendor's own results more weight as its tests add up. Finnrick's scoring settings also list a half-life of 730.5 days, about two years, so the age of a test is part of the calculation. In the Sep 25, 2026 data:
- a vendor with 1 test, which passed, shows 79%;
- vendors with 4 passes out of 4 show 82% to 84%;
- a vendor with 18 passes out of 19 shows 86%;
- a vendor with 135 passes and 27 fails, a raw pass rate of 83%, shows 72%.
The last example shows that the score isn't simple division. Across the 137 vendors on this site with a score, scores ran from 44% to 88%, with a median of 69%. Read a score together with its test count, which the site shows on peptide pages and the vendors list.
A score describes past samples, not your vial
Each test is one sample, bought at one time, from one batch. A score summarizes those samples. It can't speak for a batch that wasn't tested, a vial stored or shipped differently, or a store on another domain that uses the vendor's name. It doesn't mean a product is safe or suitable for any use. Vendors sell these compounds as research chemicals, not for human use.
Where the site shows a score
- Vendor pages and the vendors list show the vendor's overall Purity focus score.
- Peptide pages show Finnrick's score for that peptide at that vendor when one exists, otherwise the vendor's overall score, marked "overall".
- The score filters on peptide pages (60%+, 70%+, 80%+) and the "Rated 70%+" label use the same number.
Scores attach only to Finnrick-listed domains
Finnrick lists the store websites it links to each vendor. PeptidesGrid shows a vendor's score next to prices only when the store's domain is one Finnrick lists for that vendor, or when a listed domain points to it, for example by a redirect. A store on any other domain appears under its own domain name, marked "not Finnrick-listed", with no score, even if it uses the vendor's name and logo. When Finnrick lists a domain for more than one vendor and it isn't clear which runs it, the site marks it "listed for 2 vendors". A fraud warning from a named source overrides every other result, and those domains show no prices.
On Sep 25, 2026, 153 of the 185 domains the site checks were Finnrick-listed. Being on Finnrick's list is not an endorsement; it only ties a domain to the vendor the score describes. See how to check a store's domain.
Disputes and changes
PeptidesGrid doesn't change Finnrick's numbers. To dispute a score, contact Finnrick; the site picks up its changes on the next fetch. If you think the site has matched a domain to the wrong vendor, ask for a review. The method page has the full rules.
Scores and counts from Finnrick's public profile pages, fetched Sep 25, 2026. Source: Finnrick. Last reviewed Sep 25, 2026.