How we rank, in full detail.

The single most important factor: independent rigor of how a vendor's products are tested. Independent third-party labs (Janoshik, Auxilium and equivalents) score highest. In-house testing - where the vendor tests their own product and publishes the result - scores far lower because the incentive to misreport is structural.

Within third-party testing we look at methods used (HPLC for purity is the floor, mass spec for identity confirmation raises the score, endotoxin and bacterial contamination testing raises it further), test coverage (per-batch testing beats occasional spot-checks), CoA freshness(the certificate must match the current batch - CoAs years old don't count), and pre-purchase availability (CoAs accessible before a buyer commits, not only on request after).

Sub-factors within Testing Standards are not equal-weighted—independence of testing and CoA freshness dominate; method breadth (mass spec, endotoxin) and pre-purchase availability are secondary modifiers within the anchor band.

A weighted composite of community sentiment, calibrated to source quality. Not all reviews are equal - and some platforms are systematically gamed.

Highest-weighted sources are independent blind-testing programs and long-running reviewers with established track records. Below that, vetted Discord communities with active mod oversight earn medium-high weight. Reddit sits at medium weight, applied with skepticism - sudden trending with coordinated positive threads is a documented astroturfing pattern. Trustpilot sits at the bottom: vendor-solicited reviews dominate the platform and the signal is unreliable.

Track record matters on top of community signal: depth of scrutiny and independently validated consistency beat coordinated promotional bursts, because hype cycles in this market routinely collapse when scrutiny arrives.

Sub-factor weighting is explicit: corroborated blind-testing and deep reviewer dossiers move the anchored band first; vetted Discord and skeptical Reddit scraping fill corroborating detail; scraped Trust aggregates only tie-break when everything else agrees—they never rehabilitate weak evidence alone.

Logistics outcomes, scored separately from communication quality. The four sub-factors are delivery success rate (orders that arrive intact and on time across multiple regions), stealth quality (packaging that reduces customs scrutiny without misrepresenting contents), customs handling (rate of seizures across tested regions), and most critically the reship/refund policy on seizures and lost packages—which separates vendors who back their orders from vendors who offload losses onto buyers.

Customs/delivery outcomes and documented reship protections are graded as a matched pair anchoring shipping scores: heroic policies cannot paper over chronic seizures, and clean delivery stats cannot compensate for refusing documented reships—the rubric insists on credible performance across both fronts for any 3+/5 anchor.

Price evaluated relative to verified purity, never raw cost. Suspiciously cheap vendors lose points rather than gain them - in this market, lowest price strongly anti-correlates with synthesis quality, and a 99% purity product at the bottom of the price range cannot be honestly delivered.

This criterion also captures pricing transparency (clear unit pricing, no hidden fees), volume discounts (predictable bulk pricing tiers), and payment options (crypto, cards, and direct bank transfer supported).

Price relative to verified purity drives placement first—payment breadth and discount transparency tighten or loosen the verdict inside the anchored band rather than overwriting it outright.

Customer-service responsiveness, communication quality during issues, and general transparency about sourcing, manufacturing, and policies. Weighted lower than the others because it's the most subjective and the easiest area to perform well in for the wrong reasons (a vendor with great support and bad product is still a bad vendor).

Sub-factor emphasis: truthful sourcing and manufacturing disclosure plus durable issue resolution outweigh cheerfully fast replies—policy clarity and follow-through move the anchor more than tone alone.

The core question: does the app correctly do what a peptide user actually needs? That means accurate reconstitution math, correct dosing volume calculations, reliable schedule logic (every-other-day, fasted timing, cycling), and meaningful stack support. Conflict handling is split into three buckets: scheduling conflicts (doses too close together, fasted-window violations), pharmacological interactions between stacked peptides, and interactions with user-entered prescription medications. Separate from conflict handling, we evaluate unit conversion: correct conversion between mcg, mg, IU, and insulin-syringe units.

Peptide dosing crosses these units constantly, and unit errors are the most common source of dangerous miscalculation.

An app that gets the math wrong is dangerous regardless of how nice it looks. We verify every app against a fixed battery of 15–20 reference reconstitution and dosing scenarios, covering common peptides (BPC-157, semaglutide, GH secretagogues), awkward edge cases (1mg vial in 5mL water targeting 250mcg), and unit-conversion traps. The full battery and per-app results are published alongside each review.

Peptide use is sensitive health data. The privacy axis covers where data lives (local-first vs cloud-only), encryption (E2E vs in-transit only vs none), telemetry (whether the app sends usage data - and what data - to its developer or third parties), third-party SDKs (analytics, ad networks, crash reporters), account requirements (whether you can use the app at all without handing over identifying information), app-level security (biometric lock, session timeout, 2FA on cloud accounts where applicable), and server jurisdiction (where synced data physically resides and which legal regime governs it).

Open-source code isn't required for a high privacy score, but it provides a meaningful uplift here because it lets independent reviewers verify data-handling claims rather than taking the developer's word.

We audit data flows during the testing window using network inspection, not just developer claims.

Does the app actually work over time? Notification delivery rate in the test window, data persistence (no logged doses lost), sync stability (no duplicates or conflicts), offline handling (the app keeps working when the network doesn't), and data portability: CSV and PDF export, import from competitor apps, and the ability to generate a clean log to share with a clinician. Lock-in is a reliability risk—if the app shuts down, users need their history.

We test for at least 30 days per app, with a fixed protocol of scheduled doses across daypart boundaries, time zone changes, and intentional offline periods.

Design quality, accessibility (VoiceOver, TalkBack, dynamic type, sufficient contrast), onboarding clarity, and platform availability.Single-platform apps are not automatically penalized, but cross-platform parity contributes to the score because most users have a phone, a partner's phone, and a tablet or web context. A polished single-platform indie app can still score 3/5 here; it cannot reach 5/5.

Pricing model clarity, free-tier honesty, and absence of dark patterns (forced trials, hidden auto-renewal, deceptive cancellation flows). This category captures whether the business model is honest—separate from whether the code is auditable (which we score under Privacy).

Does the app surface non-obvious safety information at the points users need it? We look for sterile reconstitution technique guidance, warnings about counterfeit and untested supply, prompts to involve a clinician for symptoms or unexpected reactions, and references to third-party lab testing. Peptides occupy a regulatory gray zone in most jurisdictions—the app is often the only safety touchpoint a user encounters. Apps that ignore this, or worse, present themselves as medical advice, score poorly.