What actually makes a peptide company worth buying from?

Evidence that somebody other than the seller produced. In practice that narrows to five checks: certificates of analysis published per batch and readable without an account, the testing laboratory named rather than described as "third-party", bacterial endotoxin published as well as purity, the served markets and shipping terms stated before you order, and a direct support channel. The first two carry most of the weight, because they are the only ones a company cannot satisfy with wording alone. Everything else on a peptide site, including the design, the review count and the certifications logo strip, is easier to produce than one certificate with a laboratory name on it.

Why does the table say "not published" instead of "no"?

Because they are different statements and only one of them is true. "Not published" means we loaded that company's own site and did not find the thing there. It does not mean the company does not do it: they may test for endotoxin routinely and simply not publish the result, or publish it somewhere we did not find. Writing "no" would convert our reading of a website into a claim about a company's laboratory practice, which we have no basis for. A third label, "could not verify", covers the case where we could not read the site at all.

How was this comparison put together?

Each company's official domain was found by search rather than assumed, then loaded directly, on 29 August 2026. Each of the five criteria was recorded only against a page that actually returned content, with the exact URL and a quotation kept for each finding. Those findings were then re-checked by an independent adversarial pass instructed to refute rather than confirm, which is what caught a wrong absence claim in the first pass. Two rows were additionally corroborated by hand. Nothing here comes from a review site, a forum or a recollection: if it is not on the company's own site, it is not in the table.

Why are some cells marked "could not verify"?

Because for those we could not reach an answer we would stand behind. One site answered every request with an automated bot challenge rather than page content, so three of its five cells are unmarked: we could settle two of them from archived copies of the company's own pages, and the rest we simply do not know. Those cells are left empty rather than filled with an absence we never actually observed. Recording a blocked request as a missing certificate would manufacture a negative fact about a real business out of our own tooling, which is exactly what this comparison exists to avoid. A related trap caught us during this research: one company publishes its certificates as scanned images with no text layer, so an automated text search of them returns nothing and reads exactly like an absence. Opening them by eye showed the testing was there. Where a cell could go either way, it stays empty.

Peptara Labs is in this table. Is that not a conflict of interest?

Yes, plainly, and the honest answer is to say so rather than to hide the row. We are one of the companies being compared, we wrote the rubric, and you should weight the row accordingly. Two things limit the damage. The criteria are the same for everyone and were fixed before the checking started. And every claim in our row points at a document you can open and check yourself, which is the only defensible position for a company scoring itself: not "trust the row", but "here is the certificate, go and read it".

Does a named testing lab mean the results are correct?

No. It means the results are attributable, which is a lower bar and a necessary one. A named laboratory can be contacted, its reports carry identifiers that can be quoted back to it, and a claim about it can be checked and therefore disproved. An unnamed laboratory offers none of that. The stronger signal, where it exists, is a second report on the same batch from a different laboratory, because two independent measurements that agree closely are much harder to arrange than one report that says what a seller wants.

Comparison

Best Peptide Companies, Compared on What They Publish

Q: Is Peptide Sciences still selling peptides?

As of this check their site states that they have decided to voluntarily shut down operations and discontinue the sale of their research products, and it serves that notice in place of a catalog. The same notice states the company retains its name, brand and trademarks and does not authorise any third party to use them, so a site trading under that name is worth treating carefully. That is the company's own published statement about itself, reproduced here because it materially affects anyone searching for them, rather than an assessment of the business by us.

The short answer

There is no defensible single answer to which peptide company is best, because nobody can measure what is in a stranger's vial. What can be measured is what each company publishes about its own testing. So this page fixes five criteria first, then records what each company publishes against them, and marks everything else unknown. The rubric comes before the table on purpose: a comparison whose criteria were chosen after the winner is not a comparison.

Peptara Labs is one of the companies in the table below. We wrote the rubric, which you should weight accordingly, and every claim in our own row points at a document you can open.

The rubric, before the table

Five questions, in weight order. Each one is answerable from a company's own public site, which is the constraint that makes the exercise honest: no criterion here depends on private information, a relationship, or anyone's opinion of anyone.

Batch COAs published
Are certificates of analysis published publicly, tied to a specific batch or lot, and readable without an account?
A certificate is only evidence about the vial you received if it belongs to that batch. Public and per-batch is a much higher bar than "tested", and it is the one criterion a company cannot satisfy with wording alone.
Testing lab named
Is the third-party laboratory named, as an actual laboratory, rather than described as "third-party tested"?
A named laboratory is what makes a report checkable by somebody other than the seller. An unnamed one cannot be contacted, and the claim cannot be falsified, which is the same thing as it not being a claim.
Endotoxin published
Is bacterial endotoxin testing published, either as a result or as a stated line on the panel?
Purity is one test. Endotoxin is a separate question that purity cannot answer, and it is the test most often absent when a panel is thin, which makes it a useful indicator of how deep the testing actually goes.
Serves the US, stated
Does the company publicly state that it serves United States customers, and publish its shipping terms?
Terms you can read before ordering are terms somebody can be held to. This is the least discriminating of the five criteria and is weighted accordingly.
Support channel published
Is a direct support channel published, meaning a named email, phone number or live channel rather than a form alone?
A route to a person is what you need on the day something goes wrong. It proves little about quality on its own, which is why it sits last.

What the three labels mean

Read these before the table, because the difference between the second and the third is the difference between an honest comparison and a smear. Checked on 29 August 2026.

Published

We loaded a page on that company's own site and found the thing there, and kept the URL and a quotation for it.

Not published

We loaded their site and did not find it. This is not a claim that the company does not do it. They may test and not publish, or publish it somewhere we did not find.

Could not verify

We could not read the site at all, so we know nothing either way. A blocked request is a fact about our tooling, never evidence about a company.

How the checks were run

Each company's official domain was found by search rather than assumed, then loaded directly, and each finding was recorded with the exact URL and a quotation. Those findings were then re-checked by an independent pass instructed to refute rather than confirm, which caught a wrong absence claim the first pass had made and overturned four cells in total. Several rows were additionally corroborated by hand, and one finding that could not be reproduced on re-checking was withdrawn rather than published. Where a live site could not be read, an archived capture of that company's own pages was used instead and is labelled as such in the detail, because it is a weaker basis than a live read. Several of these brands have lookalike domains in search results, so the exact host checked is printed beside every company: if you found a brand on a different host, the row may not describe the site you are looking at.

The comparison

Listed alphabetically. There is no score, no ranking and no winner, because the criteria do not combine into one number in any way that would survive scrutiny. Read the row, then go and read the company's own documents.

Research peptide suppliers against five publicly verifiable criteria, checked on 29 August 2026
Company	Batch COAs published	Testing lab named	Endotoxin published	Serves the US, stated	Support channel published
AminoVaultaminovault.com	Published	Published	Published	Published	Published
Core Peptidescorepeptides.com	Published	Published	Not published	Published	Published
Peptara Labspeptaralabs.ioThis site	Published	Published	Published	Published	Published
Peptide Sciencespeptidesciences.com	Published	Not published	Not published	Not published	Published
Phoenix Peptidephoenixpeptide.com	Could not verify	Could not verify	Could not verify	Published	Published
Swiss Chemsswisschems.is	Published	Published	Published	Published	Published

Checked 29 August 2026. A website is a snapshot: any of these can change the day after this was written, so treat the table as a worked example of the method rather than a standing verdict, and run the five questions yourself.

What was actually found, company by company

The table compresses each finding to one word. This is the uncompressed version, so you can see what each cell rests on rather than taking the pill at face value.

AminoVault

aminovault.com

Batch COAs published: Published. A public lab-tests page links certificate PDFs that open without an account.
Testing lab named: Published. The certificates name ILS Laboratories as the testing laboratory.
Endotoxin published: Published. Certificates carry an endotoxin line citing the USP chapter 85 method.
Serves the US, stated: Published. The site states that sourcing, synthesis, testing and shipping are handled within the United States.
Support channel published: Published. A named support email address is published in the shipping and returns policy.

Every criterion was found published on their own site and confirmed on a second independent pass. One nuance worth knowing if you check this yourself: on the certificates reviewed, the lot string repeats across several products, so the per-certificate identifier to follow is the certificate or accession number rather than the lot field.

Core Peptides

corepeptides.com

Batch COAs published: Published. Product pages link certificate PDFs named for a specific lot, hosted on their own domain.
Testing lab named: Published. The certificates reviewed name Vanguard Laboratory, with a street address, a phone number, a named laboratory director and an accreditation certificate reference.
Endotoxin published: Not published. The certificate reviewed reports identity and chromatographic purity only, with no endotoxin line.
Serves the US, stated: Published. A shipping page states United States delivery terms, including a free domestic threshold.
Support channel published: Published. A contact page publishes a support email address and a telephone number.

Several near-identical domains use this brand name in search results. The row above describes corepeptides.com specifically, chosen because its published support address matches that domain and its certificate PDFs are hosted there. If you found the brand on a different host, this row may not describe the site you are looking at.

Peptara Labs

peptaralabs.ioThis site

Batch COAs published: Published. 14 released certificates, each with its own certificate number, lot and test date, downloadable in full without an account.
Testing lab named: Published. Titreon Analytical runs the full panel, with independent Janoshik endotoxin reports on select batches.
Endotoxin published: Published. Endotoxin is on the published panel, and select batches carry a second endotoxin report from a different laboratory.
Serves the US, stated: Published. The United States is a served market, with flat-rate tracked shipping and terms published before checkout.
Support channel published: Published. A direct messaging channel is published site-wide, alongside in-site chat.

This is our own site, so treat this row with the scepticism you would apply to anyone marking their own work. Every claim in it links to the document behind it, which is the only reason it belongs in the table at all: check the certificates rather than the row.

Peptide Sciences

peptidesciences.com

Batch COAs published: Published. Certificate images remain publicly reachable and their sitemap still indexes them, each carrying a lot number, but the site pages themselves now serve only the shutdown notice, so nothing is browsable from the site.
Testing lab named: Not published. No testing laboratory is named on the certificates reviewed.
Endotoxin published: Not published. No endotoxin document was found among the published certificate assets.
Serves the US, stated: Not published. Every shipping and contact page now redirects to the shutdown notice.
Support channel published: Published. The notice publishes a contact address for reporting unauthorised use of the brand.

As of this check their homepage states that the company has decided to voluntarily shut down operations and discontinue the sale of its research products, and every other page we tried served that same notice. The notice also states the company retains its name, brand and trademarks and does not authorise any third party to use them, which is worth knowing if you encounter a site trading under this name. That is the company reporting its own status, not an assessment by us. Their certificate images do remain reachable and carry lot numbers, which is why the first column reads published even though there is no longer a catalog to browse.

Phoenix Peptide

phoenixpeptide.com

Batch COAs published: Could not verify. No page could be read.
Testing lab named: Could not verify. No page could be read.
Endotoxin published: Could not verify. No page could be read.
Serves the US, stated: Published. An archived capture of their own order information page publishes shipping terms and routes European buyers elsewhere, alongside a United States headquarters address and pricing in dollars. Inferred from those together rather than from one explicit sentence, and from an archived page rather than a live one.
Support channel published: Published. An archived capture of their own contact page publishes a postal address, a telephone number and an email address. The live page could not be read from our checks.

Every request to the live site from our checks was answered by an automated bot challenge rather than by page content. That is a fact about our tooling, not about the company, so the three unfilled cells mean unchecked and must not be read as an absence. Two criteria were settled from archived captures of the company's own pages, which is a weaker basis than a live read and is labelled as such above. The remaining three stay unmarked because a partial and possibly stale archive of a large catalog cannot establish that something is absent from the current site. A browser visit from a normal connection may well load it fine, in which case the five questions in the rubric are the ones to ask.

Swiss Chems

swisschems.is

Batch COAs published: Published. Certificate PDFs hosted on their own domain identify the manufacturer and a batch.
Testing lab named: Published. Their FAQ states batches are submitted for independent analysis principally to Janoshik Analytical.
Endotoxin published: Published. A dedicated endotoxin report from the named laboratory, carrying a batch and a result in endotoxin units per vial, is linked from the product page and was read by eye.
Serves the US, stated: Published. A shipping page states United States domestic carriers and options alongside international shipping.
Support channel published: Published. A support email address and a telephone number are published on their quality page.

As with Core Peptides, several lookalike domains carry this brand name. The row describes swisschems.is, which is the host whose published support address matches the domain and whose certificate PDFs are served from it. One methodological note, because it nearly produced a wrong cell here: their certificates are published as scanned images with no text layer, so an automated text search of them returns nothing and reads exactly like an absence. The endotoxin report had to be opened and read by eye to confirm it exists, and it sits on the product page rather than in the central test-results library. A text-only check of that library would have concluded the opposite.

What this comparison cannot tell you

Worth stating plainly, because a table looks more authoritative than it is.

Whether a certificate is genuine

Publishing a document and the document being authentic are different things. That is settled by taking the identifier on the report to the laboratory that issued it, which is a check only you can run for your own vial.

What is in the vial you receive

A certificate describes the batch it was written for. It says nothing about a later batch, and a company with excellent published testing can still ship you something the certificate does not cover.

Anything about outcomes

These are analytical criteria about documents. None of them speaks to whether any compound is appropriate for any person, which is a clinical question and not a purchasing one.

The useful part

The rubric outlasts the table

Any of these sites can change tomorrow, and a comparison page written today is stale the moment one of them does. The five questions do not go stale. Take them to whoever you are actually considering, including us, and see how far you get. Our own certificates are published in full so you can start with the hardest one.

Certificates of analysis →How to verify peptide purity →Where to buy peptides →

Keep reading

The buyer checklist, the verification walkthrough, and the documents

Where to buy peptides

The buyer-side version of this rubric, and why there is no store near you.

How to verify peptide purity

A real published report read line by line, and how to check it with the lab.

Certificates of analysis

All 14 released batch certificates, downloadable in full.

How to read a COA

What each test on a certificate is actually measuring.

Trust and legitimacy

The entity behind this site, the testing, and the review record.

Are peptides legal

Research-use status and FDA standing, in general terms.

Comparing peptide companies: FAQ

General educational information only, research-use framing, not legal or medical advice. Every statement about another company describes what was published on that company's own website on 29 August 2026 and nothing more. Company names and trademarks belong to their respective owners, and their use here is for identification in a factual comparison. Websites change: verify anything that matters to you at the source before acting on it.