As of May 2026, Kovera Labs has published 1,816 COAs across 146 vendor-clients. 29 vendors have 20 or more records and qualify for a meaningful ranking. Ion Peptide leads at 105 records, zero failed purity tests, zero identity-mismatch flags, and a 99.94% median purity. Glacier Aminos (77 records) and Instant Peptides (72 records) are the only heavy publishers in the cohort with concentrated quality issues - three identity-mismatch flags each, plus five explicit purity failures at Instant Peptides. Every other vendor at 20+ records has a zero-flag record in the period.
What this leaderboard is
We mirrored every COA Kovera Labs has published as of late May 2026 - 1,816 records across 146 vendor-clients spanning January through May. Of those 146 vendors, 29 have 20 or more records and qualify for a ranking where the per-vendor signal is statistically meaningful.
The ranking is mechanical. For each vendor we count:
- Records: total Kovera COAs published in the period
- Fail: records where Kovera marked
passed: false(explicit purity failure) - ID-mis: records where Kovera marked
identityConfirmed: false - <95%: records where the measured purity came in below 95%
Vendors are ranked by clean records first (total minus all flag types), with total record volume as the tie-breaker. The structural argument: a vendor with 105 clean records is making a stronger publication commitment than a vendor with 25 clean records, even though both have zero flags.
The full leaderboard
| Rank | Records | Fail | ID-mis | <95% | Median purity | Vendor |
|---|---|---|---|---|---|---|
| 1 | 105 | 0 | 0 | 0 | 99.94% | Ion Peptide |
| 2 | 77 | 3 | 3 | 0 | 99.93% | Glacier Aminos |
| 3 | 72 | 5 | 3 | 0 | 99.92% | Instant Peptides |
| 4 | 58 | 0 | 0 | 0 | 99.94% | Zenergy |
| 5 | 54 | 0 | 0 | 0 | 99.95% | Coffee and Peppers |
| 6 | 52 | 0 | 0 | 0 | 99.93% | Mile High Compounds |
| 7 | 40 | 0 | 0 | 0 | 99.93% | Fixion Fuel |
| 8 | 40 | 0 | 0 | 0 | 99.90% | BioEdge Research Labs |
| 9 | 35 | 0 | 0 | 0 | 99.94% | Axion Research |
| 10 | 34 | 0 | 0 | 0 | 99.92% | Onyx BioLabs |
| 11 | 31 | 0 | 0 | 0 | 99.89% | Upgrade |
| 12 | 29 | 0 | 0 | 0 | 99.87% | personal-email submitter (1) |
| 13 | 27 | 0 | 0 | 0 | 99.93% | D&H Peps |
| 14 | 26 | 0 | 0 | 0 | 99.89% | Modern Aminos |
| 15 | 26 | 0 | 0 | 0 | 99.84% | GMR Compounds |
| 16 | 25 | 0 | 0 | 0 | 99.95% | SYNBIOTIX (discountpeptides.co) |
| 17 | 25 | 0 | 0 | 0 | 99.96% | Peak Chain / Genetix Chain |
| 18 | 25 | 0 | 0 | 0 | 99.82% | TriVial BioWorks |
| 19 | 25 | 0 | 0 | 0 | 99.88% | Hamner Research |
| 20 | 26 | 2 | 0 | 1 | 99.93% | Peptide Partners |
| 21 | 23 | 0 | 0 | 0 | — | Optimum Formula |
| 22 | 23 | 0 | 0 | 0 | 99.94% | personal-email submitter (2) |
| 23 | 22 | 0 | 0 | 0 | 99.88% | Renova Peptides |
| 24 | 22 | 0 | 0 | 0 | — | Beast Labs |
| 25 | 22 | 0 | 0 | 0 | 99.86% | Reta One Labs |
| 26 | 22 | 0 | 0 | 0 | 99.87% | Zenith Bioscience |
| 27 | 21 | 0 | 0 | 0 | 99.91% | REMEDYX LABS |
| 28 | 21 | 0 | 0 | 0 | — | Alpha One Aminos |
| 29 | 20 | 1 | 0 | 1 | 99.30% | Axon Peptide |
What the ranking says
Ion Peptide is decisively #1. 105 records, zero flags, median 99.94%. They publish more Kovera COAs than the next two vendors combined, and the next two vendors (Glacier Aminos and Instant Peptides) are precisely the two with concentrated quality issues. Our per-vendor Ion Peptide audit goes deeper, including the structural coded-SKU caveat the leaderboard doesn't capture.
The flag cluster is concentrated in two vendors. Of the 29 qualifying publishers, only three carry any flag: Glacier Aminos (3 identity-mismatch + 3 fails), Instant Peptides (5 fails + 3 identity-mismatch), and Axon Peptide (1 fail + 1 sub-95% purity record). Every other heavy publisher in the cohort has zero flags. That's a cleaner cohort distribution than we see in the Janoshik corpus.
The middle of the leaderboard is dense with smaller players. Below rank 5, the publishing volume tightens fast - 52 records at #6 (Mile High Compounds) drops to 22 records by #23-26. Roughly half the leaderboard sits in the 20-35 record band, which is enough volume to mean "this vendor has a routine publishing pattern" but not so much that we can draw strong batch-consistency conclusions.
Two slots are personal-email submitters. The Kovera registry lists two vendors under personal-email identities (e.g., jakestudzinski@studzpeptides.com, support@purasynth.com) rather than under their consumer-facing brand names. We list those anonymized in the public ranking. Their records are clean. Buyers who want to identify them can use Kovera's verifier directly.
Peptide Partners is a quiet warning. Two failures plus a sub-95% record at rank 20. Not the same severity as Glacier or Instant (no identity-mismatch flags), but enough that the audit watches the next 60 days of their publishing for whether the pattern persists or self-corrects.
What the leaderboard does NOT measure
Five honest limitations.
It only measures Kovera-tested behavior. A vendor that tests through Janoshik or another lab and not through Kovera does not appear on this leaderboard. The absence of a vendor here is not evidence about their actual quality - it's evidence that they don't (yet) route batches through Kovera. Janoshik's vendor-pool overlaps minimally with Kovera's.
It only measures the five-month publication window. The Kovera corpus only goes back to January 2026 because that's when Kovera began publishing. Long-term batch-to-batch consistency is not yet measurable from this dataset. A vendor with 100 clean records over five months has not yet been observed across a full year.
It does not measure what's in the vial in your hand. A vendor with 100 clean Kovera COAs can still ship a different batch than the one tested. The COA-to-vial-lot-match check is buyer-side. We cover this caveat in detail in our 7,164 Janoshik tests analysis - the same logic applies here.
Self-selection bias is real. Vendors choose which batches to submit to Kovera. A vendor that runs internal pre-tests and only submits favorable batches will look cleaner on this leaderboard than they would on a randomized sample. The 0.8% explicit-failure rate in the underlying Kovera corpus is a floor, not a ceiling.
Coded SKUs introduce comparison friction. Ion Peptide ships some products under coded names (ION-3R, ION-2T, §-31) which limit how directly a buyer can compare a Kovera record to a labeled vial. We cover this in their audit profile and in our code-name catalogs article.
What this means for buyers
Three operating rules consistent with the leaderboard data.
Treat Kovera-record volume as a positive signal, not a verdict. A vendor with 50+ clean Kovera records has earned a stronger trust signal than a vendor with one COA on one batch. But the leaderboard is one of several signals you should weight, not a single ranking that overrides everything else.
Pair the leaderboard with the per-vendor audit page. Vendors that have a vialaudit profile (Ion Peptide, Modern Aminos, plus those in our /vendors index) have additional editorial context the leaderboard alone cannot carry - regulatory posture, alias-cluster checks, payment methods, shipping data, named-prescriber transparency. Use both.
Flagged vendors deserve the binary read. A vendor with one explicit purity failure is in a different category than a vendor with multiple identity-mismatch flags. Glacier Aminos and Instant Peptides have published evidence that the lab could not confirm identity on real-named single-molecule peptides (NAD+, Tesamorelin, GLP-3 RT). That is the strongest negative signal in the dataset and the one we weight most heavily in any per-vendor read.
Methodology and updates
This leaderboard is a snapshot of the Kovera Labs public verifier as of May 26, 2026. We re-pull the data weekly and refresh the ranking when the structural picture changes - new heavy publishers, new flag clusters, or vendors falling below the 20-record threshold.
The underlying mirror, the extraction pipeline, and the analytical method are described in our 1,816-COA analysis. The four-question framework we use to evaluate Kovera as a credible third-party lab in the first place is in our Janoshik legitimacy article, with the Kovera-specific section appended.
Related reading
- What 1,816 Kovera Labs COAs reveal about the US peptide market - the parent analysis this leaderboard is extracted from.
- Is Janoshik testing legit? - the four-question lab-credibility framework applied to both Janoshik and Kovera.
- How peptide vendors fail: a taxonomy - the nine failure modes the leaderboard surface-level data can and cannot detect.
- Code-name catalogs: FDA-evasion analysis - on coded SKUs like
ION-3Rand what they mean for buyer-side comparison. - Our vendor index - per-vendor audit profiles for the vendors on the leaderboard that we cover directly.
Sources
- Kovera Labs public verifier - the source-of-truth verifier audited for this piece
- Internal: 1,816-record extraction corpus at
data/kovera/(May 26, 2026 mirror) - Internal: What 1,816 Kovera Labs COAs reveal
- Internal: Is Janoshik testing legit?
Frequently asked
What is the Kovera Labs vendor leaderboard?
A ranked table of every research-peptide vendor with at least 20 published COAs in Kovera Labs' public verifier as of May 26, 2026. We rank by clean-record count - total records minus explicit purity failures minus identity-mismatch flags minus sub-95% purity records. Vendors with the same clean count are tie-broken by total record volume.
Which vendor has the most Kovera COAs?
Ion Peptide leads with 105 published Kovera records covering March to May 2026. The next closest heavy publishers are Glacier Aminos (77), Instant Peptides (72), Zenergy (58), Coffee and Peppers (54), and Mile High Compounds (52).
Why does a higher Kovera record count matter?
A vendor with one Kovera COA on one batch is publishing one data point. A vendor with 50 Kovera COAs across multiple products and dates is routinely submitting to a third-party lab that publishes failures, and the failures they have published are within the industry baseline. Multi-record publishing history is the structural signal the leaderboard rewards.
Are vendors at the top of the leaderboard automatically the safest to buy from?
No. Heavy COA publishing on Kovera is a meaningful trust signal but it is not equivalent to a vialaudit full audit. The leaderboard measures publication discipline and surface-level chemistry. It does not measure first-hand vial-in-hand testing, batch-to-batch consistency over a 12-month window, customer-service resolution, or alias-cluster patterns. Buyers should pair the leaderboard with the per-vendor audit pages on this site and with their own due diligence.
How does the leaderboard handle personal-email vendor identities?
Two vendors on the underlying Kovera registry submit under personal-email identities rather than corporate-brand identities. We list them anonymized as "personal-email submitter" in the ranking. They are technically public but unnecessary republication of an individual email is doxxing-adjacent. Buyers who want to identify those specific records can look them up directly on Kovera's verifier.