What we reviewed
We researched all 50 directory entries on 10 September 2026 using official supplier pages and selected accessible analytical records. Reviews distinguish inspected batch results, laboratory-hosted records, supplier statements, optional services and inaccessible evidence. This is a public-documentation review, not a procurement audit or physical product-testing programme. Compare within the same business type: an institutional manufacturer may provide comprehensive private customer records that are not publicly accessible.
Five dimensions. 100 possible points.
Purity & identity
Chromatographic purity with method and molecular identity. A headline percentage alone earns no points.
Independent laboratory
A named external laboratory on an accessible report. Vendor-hosted reports are distinguished from laboratory-authenticated reports.
Contaminant controls
Six equally weighted checks: endotoxins, bioburden, sterility, residual solvents, elemental impurities, and peptide-related impurity characterization.
Quantity & traceability
Quantified content and identifiable, dated sample or batch records.
Accountability
Clear operator and research-use context, plus accessible contact or policy information.
Explicit scoring rules
Purity and identity /20: 0 for unsupported percentages or general quality claims; 5 for a named purity-and-identity workflow or partial analytical documentation; 10 for an inspected relevant sample result; 15 for multiple dated samples with results. Full 20 requires representative current-lot coverage, validated methods and independently corroborated sample traceability. A high HPLC number does not automatically earn more points.
Independent laboratory /20: 0 for a claim without an inspected external report, or for in-house QC; 10 for an inspected supplier-hosted report naming an external laboratory; 15 for an inspected record on that laboratory’s own portal. Full 20 additionally requires verified accreditation scope, independence and sample custody. A laboratory portal confirms a record exists; it does not prove which delivered vial was tested.
Contaminant controls /30: six categories worth 5 each: endotoxins, bioburden, sterility, residual solvents, elemental impurities and related-peptide impurity characterization. Each earns 0 for marketing, optional service capability or an unresolved result; 2.5 for an inspected relevant sample result; 3.75 for results across multiple dated samples; 5 requires representative, independently corroborated current-lot evidence and validated methods/limits. A chromatographic main-peak percentage is not a full related-impurity characterization. Different tests do not substitute for one another.
Quantity and traceability /20: two components worth 10 each. Quantity earns 5 for an inspected content result and 7.5 for multiple dated samples. Traceability earns 5 for an identifiable lot with a dated record and 7.5 for multiple dated lots. Each component reaches 10 only with representative current coverage and independent sample-to-delivery corroboration. An undated run or an unknown batch does not earn traceability credit; a label amount is not a measured content result.
Accountability /10: 0 where an operator is not established; 2.5 for limited identity/context evidence; 5 for public business/contact information with product-use or policy context. Higher tiers require independently checked operator details, documented corrective actions and audit evidence. Company size, price, popularity and testimonials do not contribute.
Testing frequency: claims versus observations
Every-batch policies are labelled as supplier statements unless a complete batch log has been audited. A small sample of reports cannot verify universal compliance. We record grade exceptions, optional tests, sample dates and unknown intervals. Publication dates, expiry dates, annual quality reviews and weekly project updates are not automatically analysis or retest dates.
The calculation and changes
Rounding occurs once at the total. Unknowns remain in the denominator. Ties are alphabetical. Archived or insufficiently identifiable/accessed listings are unscored. Version 0.2 replaces the sparse initial assessment with supplier-specific findings and a more explicit rubric; changes can reflect better documentation or clarified rules, not a change in product quality. The open dataset retains the previous score.
What changed in the evidence
Particle’s newer July and September records support additional microbial, endotoxin and elemental documentation credit; the older GHRP-2 discrepancy remains unresolved. Biotech’s MZ Biolabs certificate adds dated purity/identity evidence. Peptx’s Janoshik portal record supports provenance, but its batch is unknown. UK Peptides explicitly describes its COAs as non-batch-specific. JPT discloses different sampling and microbial panels by grade.
Purification must leave evidence
Multiple purification rounds, high manufacturing cost, salt exchange or a pharmaceutical-grade claim do not prove removal of specific contaminants. We prioritize final-sample results, method scope, uncertainty and limits. A stated test capability earns no contaminant result points. A report may document a concerning result and still earn transparency credit; the score is not a pass/fail safety judgment.
Limits and corrections
No score reaches full maturity in this edition. No blind purchasing, independent sampling, complete batch audit or direct laboratory-staff confirmation was performed. Missing evidence does not prove contamination. Different criteria may refer to different samples; the total is a profile-level index, not one combined batch-release decision. Community ratings are separate and never affect editorial scores.
Submit a sourced correction or inspect the open dataset. Changes should identify the product, lot, laboratory, method and exact result rather than only a marketing claim.