Koi Peptides

Listings clearly state products are for laboratory research use only.

No therapeutic, dosing, or human-use claims anywhere in the listing.

Batch-tied COAs (HPLC purity + mass-spec identity) exist and are accessible.

Vial sizes and prices are listed clearly and consistently, no bait pricing.

A stated shipping, tracking, and reship/refund policy is published.

No unresolved counterfeit, non-delivery, or COA-forgery reports on file. 2026-08-26: ALL THREE COA SELF-CONTRADICTIONS CONFIRMED UNREMEDIATED, SECOND CONSECUTIVE FIRST-HAND RUN. The maker recommended coa_documentation FAIL; HELD AT PENDING and escalated to Frank - the prior run's reasoning ('the PDFs are published and lot-tied, they just weren't opened') still technically holds, but it is wearing thin and Frank should adjudicate. VERBATIM: /coa/ states 'Lot IDs follow the format KOI-YYYY-MMDD-x and are printed directly on the vial' - yet EVERY ONE of the 15 published lots is a bare 10-digit number (BPC-157 2604060200, TB-500 2512290134, GHK-Cu 2605290222, Semax 2604030181). The homepage separately advertises latest lot 'KOI-2026-0418-A - BPC-157 Acetate', a lot ID in the claimed format that APPEARS NOWHERE IN THE LIBRARY, which lists BPC-157 as 2604060200. The methods block states 'Lots must measure >=99% by integrated peak area for release' - yet Semax lot 2604030181 is published at 98.957%, a RELEASED LOT BELOW THE VENDOR'S OWN RELEASE SPEC. Coverage: 'Showing 1-8 of 15 documents' against 24 distinct catalog products; GLP-1 SG, GLP-2 TZ, Epithalon, Melanotan-II, Sermorelin, AOD-9604, Ipamorelin and PT-141 have no COA on file. RUO framing strong and consistent ('All Koi peptides are sold strictly for laboratory research use only not for human or animal consumption'); no dosing or human-outcome language found; per-vial pricing clear ($34.95-$179.95). NEW AND SHARPER THAN YESTERDAY: Koi advertises COAs 're-run by an independent ISO/IEC 17025 lab.' Read first-hand today, Freedom Diagnostics' own site claims ONLY 'Compliant with ISO 9001 Quality Management Principles' and 'GLP Quality Standard Principles' - self-declared compliance language, not accreditation - and NO ISO 17025 claim appears anywhere on it. IF Koi's lab is Freedom Diagnostics (suggested by the GHK-Cu COA filename ...-fd0618.pdf, which is filename inference and NOT verified), then Koi's ISO/IEC 17025 claim is false on the lab's own published position. That conditional must be resolved before it is published - identify Koi's lab first. The Blank-Peptides->Koi COA-alteration allegation remains UNVERIFIED (Trustpilot empty body, third consecutive day); it is the single highest-severity open item on this vendor and it is aging without resolution - recommend a browser read. || PRIOR: 2026-08-25: TWO NEW FIRST-HAND SELF-CONTRADICTIONS IN THE COA LIBRARY, both found by reading the vendor's own stated rules against its own published data. (1) LOT FORMAT MISMATCH: the page states verbatim 'Lot IDs follow the format KOI-YYYY-MMDD-x' and the homepage advertises latest lot 'KOI-2026-0418-A' - but EVERY actual lot ID in the table is a bare 10-digit number (2604060200, 2512290134, ...). The advertised format matches NOTHING on file. (2) BELOW ITS OWN RELEASE SPEC: the methods block states 'Lots must measure >=99% by integrated peak area for release', yet Semax lot 2604030181 is published at 98.957% - below the threshold the vendor set for itself. Neither is a health claim or a forgery; both are accuracy defects in the exact artifact the coa_documentation check inspects, and together they weaken confidence in the library's internal discipline. coa_documentation HOLDS AT PENDING rather than moving to fail: the PDFs ARE published and ARE lot-tied, and they were not opened this run, so this is a documentation-integrity concern, not a finding that testing is absent. Coverage is still 15 documents against ~24 catalog SKUs - GLP-1 SG, GLP-2 TZ, PT-141, Ipamorelin, Sermorelin, AOD-9604, Melanotan-II, Epithalon and the Wolverine blend have no COA row. RUO FRAMING IS THE STRONGEST OF TODAY'S BATCH AND SHOULD BE CREDITED: 21+ interstitial, footer 503A/503B disclaimer, and 'For Research Use Only - Not for human or veterinary use' on every page. no_scam_flags STAYS PENDING, unchanged: the Blank Peptides -> Koi Peptides COA-alteration allegation remains UNVERIFIED (Trustpilot bot-blocked two days running) and moves nothing until read first-hand. RELATED AND NOW SHARPER: Koi advertises COAs 're-run by an independent ISO/IEC 17025 lab'. A direct search of Freedom Diagnostics - suggested here by the GHK-Cu COA filename coa-ghk-cu-100mg-fd0618.pdf, though that is FILENAME INFERENCE, NOT VERIFIED - found NO ISO/IEC 17025 accreditation, only a reported CLIA registration (SECONDARY, unverified). CLIA and ISO 17025 are not equivalent. If Koi's 17025 claim rests on Freedom Diagnostics, the claim is inaccurate. Recorded not scored pending verification of which lab Koi actually means. || PRIOR: 2026-08-24: OPEN QUESTION RESOLVED FIRST-HAND: homepage DOES list 'GLP-1 SG 10mg' linking to /product/semaglutide/ — the listing is live today; the 08-20 'correction' is the version that was wrong. /catalog/ loaded normally over HTTPS (yesterday's SSL failure resolved). /coa/ still 15 lot-tied docs; live SKUs uncovered incl. GLP-1 SG, GLP 2(TZ), Melanotan-II, MOTS-c, Epithalon, PT-141, Ipamorelin, Sermorelin, AOD-9604 — coa stays pending. NEW: no_scam_flags -> PENDING on an UNVERIFIED search-surfaced Trustpilot review alleging a counterfeit COA (client name 'Blank Peptides' allegedly altered to 'Koi Peptides') and noting Freedom Diagnostics is not ISO 17025-accredited despite Koi's 'ISO/IEC 17025 lab' claim. Direct Trustpilot fetch returned empty (bot-blocked) — rests on search summary ONLY; needs manual verification before it can move anything further. RUO framing strong (age-gate, FDA disclaimer, per-page RUO). || PRIOR: 2026-08-23: /coa/ RE-READ: now 15 lot-tied documents with HPLC purity, LC-MS identity and endotoxin — real growth — but coverage is still partial: no COA rows for live SKUs incl. Melanotan-II, Ipamorelin, Sermorelin, AOD-9604, PT-141, Epithalon; coa STAYS PENDING. Lot-ID format internally inconsistent (page says KOI-YYYY-MMDD-x, table uses numeric IDs like 2604060200). PRIOR-STATE CONTRADICTION, RECORDED FOR RE-VERIFICATION NOT ADJUDICATED: the homepage today lists 'GLP-1 SG 10mg' at /product/semaglutide/ alongside GLP 3 (Reta) and GLP 2 (TZ) — this contradicts the 2026-08-20 correction that semaglutide is not a Koi SKU. One of the two reads is wrong; re-verify /catalog/ next run before relying on either. /catalog/ fetch FAILED this run (ERR_SSL_BAD_RECORD_MAC_ALERT), so the contradiction could not be resolved today. || PRIOR: 2026-08-20: /methodology/ AND /catalog/ BOTH OPENED FIRST-HAND - EVERY PAGE IN THE KOI THREAD HAS NOW BEEN READ. THE CORRECTIONS COME FIRST BECAUSE THEY CUT FOR THE VENDOR. CORRECTION 1: SEMAGLUTIDE IS NOT A KOI SKU. Yesterday listed 'GLP-1 SG / semaglutide' among the nine uncovered SKUs and called the two uncovered GLP-1-class products the sharpest point in the finding. THE CATALOG CARRIES EXACTLY TWO GLP-CLASS PRODUCTS: GLP-3 (Reta) and GLP-2 (TZ). There is no semaglutide product page. RETATRUTIDE IS COVERED (lot 2606170094, 99.76%). ONLY TIRZEPATIDE IS UNCOVERED. The point stands at half its former weight and the correction is recorded ahead of everything else, because a finding that overstated a vendor's exposure is the failure mode this review exists to avoid. Residual, recorded not scored: the /catalog/ meta description still reads 'Shop BPC-157, semaglutide, TB-500, retatrutide and more' - a stale SEO tag advertising a compound not in the catalog; not a compliance finding, noted so it is not later read as a hidden SKU. CORRECTION 2: KISSPEPTIN-10 AND BACTERIOSTATIC WATER ARE LIVE CATALOG SKUs. Yesterday recorded them as 'two library entries that do not appear in the catalog'. Both are on /catalog/ (Kisspeptin-10 $54.95-$89.95; bac water $14.95-$19.95 under 'Research Supplies'). That observation was an artefact of counting against the homepage. RETRACTED. THIS ALSO SETTLES THE BAC-WATER QUESTION AND KOI MUST COME OUT OF THE 'AFFIRMATIVELY REFUSES' COLUMN. Koi's /about-us/ sentence - 'We don't sell peptides BUNDLED with bacteriostatic water or syringes' - is accurate as written and is about bundling. Koi does sell bac water as a separate SKU. Its placement in the 'refuses' column (made 2026-08-18) is wrong and is removed here. UNDER THE DOSAGE-FORM RULE PROPOSED 2026-08-17 KOI STILL PASSES: the full 28-SKU catalog contains no syringes, no alcohol prep pads, no sprays, strips or tablets. Same outcome as Precision Research Peptides, on the same reasoning. THE COVERAGE GAP, RECOUNTED AGAINST THE AUTHORITATIVE PAGE. The catalog holds 28 DISTINCT SKUs, not ~24. Fifteen are directly lot-matched to the fifteen published certificates. Four are multi-vial 'Stacks' whose components are individually covered - counted as covered, a judgement call made IN THE VENDOR'S FAVOUR. NINE SINGLE-VIAL SKUs HAVE NO PUBLISHED CERTIFICATE: GLP-2 (TZ)/tirzepatide, AOD-9604, Wolverine co-lyophilised blend (components covered individually, the blend is not), standalone Ipamorelin, Sermorelin, MOTS-C, Epithalon, PT-141, Melanotan-II. NINE AGAIN BUT NOT THE SAME NINE - semaglutide is out, MOTS-C is in and was missed yesterday. Counted strictly (stacks excluded) the gap is thirteen. coa_documentation HOLDS AT pending. It is a documentation gap, NOT an allegation of fraud, and it reverses the moment the nine certificates are published. FAVOURABLE AND MATERIAL: the site carries a 'COA Coming Soon - the latest batch is currently being tested' notice with email capture. Koi acknowledges pending certificates rather than presenting the library as complete. That is the honest handling of a coverage gap and it weighs against the homepage's absolute 'Every lot's COA is published on its product page.' YESTERDAY'S FREEDOM DIAGNOSTICS ACCESSION INFERENCE IS NOW CONFIRMED BY KOI'S OWN DOCUMENTATION. /methodology/ Stage 7, verbatim: FreedomDiagnostics 'issues the analytical results that appear on the Koi COA under its own letterhead, WITH A UNIQUE VERIFICATION ID PER SUBMISSION', and the buyer-documentation list includes 'THE FREEDOMDIAGNOSTICS VERIFICATION ID THAT CAN BE CHECKED AGAINST FREEDOMDIAGNOSTICS'S OWN RECORDS'. The ten-digit strings in the library are the LABORATORY'S identifier, not a Koi manufacturing lot. This resolves Frank's open item on the portal lookups without needing the portal. BUT THE SAME PAGE STATES THE OPPOSITE OF WHAT THE LIBRARY SHOWS: 'documentation that ties back to a single lot number, WHICH IS PRINTED ON THE VIAL AND INDEXED IN THE COA LIBRARY. Buyers can pull the full analytical record for any lot they receive by searching that lot number.' /coa/ says that string looks like KOI-YYYY-MMDD-x. All fifteen library entries are ten-digit numbers. The methodology page names two distinct identifiers and the library appears to be indexed by the second while labelled as the first. A BUYER FOLLOWING KOI'S OWN WRITTEN INSTRUCTIONS STILL CANNOT MATCH THE VIAL TO THE LIBRARY. Unchanged in force from yesterday, now supported by the vendor's own text rather than cross-vendor inference. RECORDED NOT SCORED - a labelling defect, not a chemistry defect. Neither Koi nor the lab has confirmed it; well-supported, not established fact. THE PURITY SPEC IS NOW A THREE-WAY CONFLICT AND IT MOVED AGAINST THE MARKETING COPY, NOT FOR IT. /quality/ says >=98.0%; /coa/ says >=99%; /methodology/ NEW says 'Acceptance criterion: >=99% for the standard Koi catalog'; footer badge is hedged '>=99% Purity Target'. SEMAX LOT 2604030181 IS PUBLISHED AT 98.957% AND IS IN THE RELEASED LIBRARY - it now passes under ONE PAGE IN THREE, not one in two. Yesterday's framing ('the quality documentation is more conservative than the marketing and the quality documentation matches the evidence') NO LONGER FITS: /methodology/ is a quality document and it says >=99%. This is closer to a genuine internal specification conflict than a copy defect, and it is recorded that way because the earlier, more charitable reading is not supported. Publishing 98.957% rather than rounding remains honest reporting and that credit stands. NO CHECK MOVES - a single lot, 0.043 points under one of three stated criteria, no allegation attached. ISO/IEC 17025 TALLY UPDATED: FOUR SURFACES ASSERT IT (homepage hero, FAQ, /about-us/, /coa/), THREE NAME THE LAB AND DECLINE THE CLAIM (/quality/, /methodology/ NEW, and Koi's own blog). /methodology/ is the third page to name FreedomDiagnostics as sole laboratory with no accreditation claim WHILE VOLUNTEERING FIVE LIMITS: not a compounding pharmacy (503A/503B), NOT A cGMP MANUFACTURER, analytical program 'not validated to the standard required for human therapeutic use', DOES NOT TEST FOR STERILITY (USP <71>), no viral clearance or mycoplasma testing. A PAGE THAT VOLUNTEERS 'WE ARE NOT cGMP' IS NOT A PAGE THAT WOULD OMIT AN ACCREDITATION IT HELD. FIFTH SURFACE, SELF-INFLICTED: Koi's own blog article 'HPLC vs Mass Spectrometry' - a standard-setting piece - lists among the marks of a defensible COA 'ISO/IEC 17025-accredited release records' and 'third-party release: ISO/IEC 17025-accredited testing where possible'. KOI IS PUBLISHING A SUPPLIER-EVALUATION STANDARD ITS OWN PUBLISHED LAB DOCUMENTATION DOES NOT CLAIM TO MEET. The 'where possible' hedge does real work and the article is otherwise clean - rigorous analytical chemistry, correctly cited, ZERO HEALTH CLAIMS, no dosing language. THIS FILE CONTINUES TO ASSERT NOTHING about whether FreedomDiagnostics holds the accreditation; absence of a claim is not proof of absence. FOUR LISTED VENDORS REST ON THIS LAB and the question is answerable in one email. /methodology/ IS THE MOST CANDID PROCESS DOCUMENT READ IN THIS DIRECTORY: eight stages from raw-material sourcing through SPPS, cleavage, preparative HPLC, counter-ion handling, lyophilisation, QC and shipment, with named reagents (HATU, HBTU, DIC, Oxyma, DIPEA), named resins (Wang, Rink amide, 2-chlorotrityl), in-process Kaiser/chloranil monitoring, ICH Q3C solvent classes, USP <85> endotoxin, ICP-MS heavy metals, and peer-reviewed citations. Nothing comparable exists on any other vendor site in this dataset. RECORDED NOT SCORED: /methodology/ and the blog are authored or reviewed by Dr. Tshering Pedon, MBBS. The content read is analytical chemistry with no therapeutic or dosing language anywhere, but medical-credential signalling on a research-use vendor's byline is worth a reviewer's awareness. It moves nothing. NO COA PDF HAS BEEN OPENED - coverage and traceability verified, analytical content NOT verified. STATUS HOLDS AT monitoring. ALL SIX CHECKS UNCHANGED. NOT A FLAG AND NOT A FLAG CANDIDATE. || PRIOR: 2026-08-19: /coa/ WAS OPENED FIRST-HAND - THE LAST OUTSTANDING GAP FROM 2026-08-18 IS NOW CLOSED, AND IT CUTS BOTH WAYS. THE LIBRARY IS REAL: 15 certificates, each with a lot ID, an HPLC purity value to three decimals, a tested date and a linked PDF (BPC-157 2604060200 99.283% Apr 7 2026; TB-500 2512290134 99.127%; GHK-Cu 2605290222 99.96%; KLOW 2604060227 99.789%; GLOW 2605290225 99.96%; Tesamorelin 2512290135 99.059%; SS-31 2603200300 99.924%; KPV 2602060018 99.899%; Kisspeptin-10 2603200302 99.256%; Selank 2602060019 99.876%; Semax 2604030181 98.957%; NAD+ 2512290139 99.842%; CJC-1295+Ipamorelin 2604060224 99.928%; Retatrutide 2606170094 99.76% Jun 20 2026; Bacteriostatic Water 2604030180 99.964%). That is a genuine, inspectable corpus and it is more than most vendors in this directory publish. FINDING 1 - COVERAGE GAP, QUANTIFIED, AND IT IS THE OPERATIVE BASIS FOR HOLDING coa_documentation AT pending. The homepage states 'Every lot's COA is published on its product page.' The catalog rendered on the homepage carries roughly 24 distinct SKUs. THE LIBRARY COVERS 15 DOCUMENTS AND NINE LIVE SKUs HAVE NO PUBLISHED CERTIFICATE: GLP-2 (TZ)/tirzepatide, GLP-1 SG/semaglutide, Epithalon, Melanotan-II, PT-141, Ipamorelin, Sermorelin, AOD-9604, and the Wolverine BPC-157+TB-500 blend (the two components are covered individually, the blend is not). MOST POINTEDLY, THE TWO GLP-1-CLASS SKUs - SEMAGLUTIDE AND TIRZEPATIDE - ARE AMONG THE UNCOVERED, AND THOSE ARE PRECISELY THE COMPOUNDS AT THE CENTRE OF 2025-2026 FDA ENFORCEMENT. Compare directly with Precision Research Peptides, verified 2026-08-18 at 16 of 16 live SKUs lot-matched. coa_documentation HOLDS AT pending. This is a documentation gap, NOT an allegation of fraud, and it is reversible the moment the nine certificates are published. FINDING 2 - THE STATED LOT-ID FORMAT DOES NOT MATCH A SINGLE LOT IN THE LIBRARY. /coa/ states verbatim: 'Lot IDs follow the format KOI-YYYY-MMDD-x and are printed directly on the vial.' The homepage 'latest released lot' reads KOI-2026-0418-A, in that format. BUT EVERY ONE OF THE 15 LOT IDs IN THE ACTUAL TABLE IS A TEN-DIGIT NUMERIC STRING. Zero of 15 match the documented format. STRONG CROSS-VENDOR INFERENCE, RECORDED AS INFERENCE NOT FACT: those ten-digit strings are the FREEDOM DIAGNOSTICS ACCESSION-NUMBER FORMAT - the Protide Health certificate read first-hand on 2026-08-16 carried accession 2608070002 in exactly the same shape. If correct, what Koi publishes as its 'lot ID' is the LABORATORY'S accession number, not a Koi manufacturing lot. THAT IS NOT BAD - it makes each entry independently checkable in the Freedom Diagnostics public portal, which is better than an unverifiable internal code - but it means the vial-printed KOI-format identifier and the published identifier are two different things, and a buyer following the site's own instructions cannot match them. THIS ALSO ANSWERS AN OPEN ITEM FOR FRANK: accession 2606170094, one of the three he was asked to look up, IS KOI'S RETATRUTIDE LOT. FINDING 3 - A PUBLISHED LOT SITS BELOW THE RELEASE SPEC THE SAME PAGE STATES. /coa/ states: 'Lots must measure >=99% by integrated peak area for release.' SEMAX LOT 2604030181 IS PUBLISHED AT 98.957%. It is in the released library. Under the >=98.0% specification written on /quality/ that lot passes cleanly; under the >=99% criterion stated on /coa/ and the >=99% footer badge, it does not. RECORDED, NOT SCORED - no check moves on this. AND THE FAIRNESS POINT MATTERS: publishing a 98.957% result rather than rounding it to 99% is HONEST REPORTING, and honest reporting is what this directory exists to reward. The defect is in the marketing copy, not the chemistry. THE PATTERN IS NOW THE FINDING, AND IT IS CONSISTENT ACROSS TWO INDEPENDENT AXES. On accreditation, /quality/ will only say independent labs 'typically' hold ISO/IEC 17025 while the homepage, FAQ, /about-us/ and now /coa/ all assert it flatly ('re-run by an independent ISO 17025-accredited laboratory', /coa/, verbatim). On purity, /quality/ commits to >=98.0% while the marketing commits to >=99% and a released lot lands between them. IN BOTH CASES KOI'S CAREFUL QUALITY DOCUMENTATION IS MORE CONSERVATIVE THAN ITS MARKETING PAGES, AND THE QUALITY DOCUMENTATION IS THE ONE THAT MATCHES THE EVIDENCE. That is a marketing-copy problem with a straightforward fix, and it should be put to the vendor as one. FAVOURABLE AND NEW THIS RUN: the homepage has moved to hedged language in several places - '>=99% purity target - the exact value for your lot is stated on that lot's COA' - which is the correct formulation and partially resolves the spec-versus-marketing mismatch recorded on 2026-08-18. The /coa/ methods section states real analytical detail (RP-HPLC integrated peak area, ESI-LC-MS identity and fragmentation, ICP-MS for Pb/As/Cd/Hg below 10 ppm, LAL endotoxin, counterion acetate/TFA reported per lot) and the homepage states a numeric endotoxin release spec of <0.5 EU/mg. The 21+ RUO interstitial renders verbatim, the footer carries complete 503A/503B non-compounder language, and the FDA disclaimer is present on both pages read. QUALIFICATION AGAINST YESTERDAY'S FAVOURABLE READ, RECORDED BECAUSE IT CUTS AGAINST KOI: on 2026-08-18 Koi was placed in the 'affirmatively refuses' column of the dosage-form prong on the strength of /about-us/ saying 'We don't sell peptides bundled with bacteriostatic water or syringes.' BACTERIOSTATIC WATER LOT 2604030180 IS IN THE COA LIBRARY AT 99.964%. Koi's sentence is about BUNDLING, not about selling, so this is not a contradiction - but a vendor that tests bac water is plausibly stocking it, and bac water was not visible in the homepage catalog either way. KOI'S PLACEMENT IN THE 'REFUSES' COLUMN SHOULD BE TREATED AS UNCONFIRMED UNTIL /catalog/ IS SWEPT. Recorded so the favourable read is not carried forward unexamined. Standing items unchanged: no second laboratory exists (settled 2026-08-18); three BPC-157 lot identities; Sheridan WY registered-agent address; PT-141 noindexed; the 'COA Coming Soon' email-capture block confirmed again today as site-wide template boilerplate, appearing on /coa/ itself. STATUS HOLDS AT monitoring. ALL SIX CHECKS UNCHANGED. NOT A FLAG, NOT A FLAG CANDIDATE. || PRIOR: 2026-08-18: THE GAP FLAGGED YESTERDAY WAS CLOSED - /quality/ AND /about-us/ READ IN FULL FIRST-HAND. THE 17025 QUESTION IS NOW ANSWERED, AND THE ANSWER GOES AGAINST THE PRODUCT-PAGE CLAIM. Yesterday's demotion was recorded with an explicit gap: /coa/, /quality/ and /methodology/ had not been opened and either could name a second, unnamed ISO/IEC 17025 laboratory performing the re-run. /quality/ is the page Koi's own 'Independent ISO/IEC 17025-accredited lab' link on /about-us/ POINTS TO. IT NAMES NO SECOND LAB. It describes a SINGLE-LAB model in unambiguous terms: 'Every analytical method that appears on a Koi COA is run at FreedomDiagnostics, an independent third-party analytical laboratory' and 'Samples are sent from Koi's synthesis facility to FreedomDiagnostics.' THE SECOND-LAB HYPOTHESIS IS NOT SUPPORTED. AND THE DECISIVE DETAIL: /quality/ ITSELF NEVER ASSERTS THAT FREEDOMDIAGNOSTICS HOLDS ISO/IEC 17025. Its only mention of the standard is generic and hedged, verbatim: 'Independent analytical labs operating in this space TYPICALLY work under ISO/IEC 17025:2017' and 'ASKING a third-party lab about its 17025 status is the standard due-diligence question for anyone evaluating analytical credibility, and the answer should be a documented one.' That is a sector generality plus an instruction to ask - NOT a claim about Koi's own lab. So Koi's most carefully written and formally reviewed quality document (Last reviewed May 14 2026, Reviewed by Dr. Tshering Pedon) STOPS SHORT of the exact claim that its product pages, its /about-us/ page and its FAQ all make flatly ('an independent ISO/IEC 17025-accredited laboratory'). This is an internal inconsistency within Koi's own site, and the more rigorous page is the one that declines the claim. coa_documentation THEREFORE HOLDS AT PENDING and status HOLDS AT MONITORING. The finding is sharpened, not reversed - but it is still NOT a fail and still NOT an accusation, and it remains reversible on a single vendor reply. NOW THE SUBSTANTIAL MITIGATION, WHICH MUST BE READ ALONGSIDE IT. /quality/ IS THE STRONGEST QUALITY DOCUMENT ANY VENDOR IN THIS DIRECTORY HAS PUBLISHED, and it is not close. It sets out: written numerical specifications fixed BEFORE testing ('The specification is not adjusted to match the result'); a full six-parameter acceptance table with methods and criteria (ESI-MS identity within 1 Da; RP-HPLC purity GTE98.0 percent area; net peptide content reported numerically; USP <85> endotoxin in EU/mg; ICH Q3C residual solvents; ICP-MS heavy metals in ppb); compendial-method framing against USP <1225> validation and USP <1226> verification; EXPLICIT DISCLOSURE OF THE FINANCIAL RELATIONSHIP with the lab ('Koi pays FreedomDiagnost ...[EARLIER NOTE HISTORY TRUNCATED IN THIS FILE; FULL CHAIN PRESERVED IN compliance-reports/]