Skip to content

How we score

Every number, and where it comes from.

Four pillars, fixed weights, no manual adjustments. Each score on a vendor page can be rebuilt from the inputs printed beside it and the nine chapters below, which also say which inputs are measured, which are estimated, and which are placeholders.

Pillars
4
Checks
7
Snapshot
Oct 28
  1. 01 The standard Three measurements, all or nothing.
  2. 02 Delivery score Days to the door, as points.
  3. 03 Price score Dollars per milligram against the median.
  4. 04 Purity score Recent passes, weighted by age.
  5. 05 Trust score Seven checks, 100 points.
  6. 06 Overall score Four pillars, fixed weights.
  7. 07 Rules of the board Who is listed, what is left out.
  8. 08 Worked example Solution Peptides, rebuilt.
  9. 09 Data status Real, estimated, placeholder.

01

The standard

A sample is judged pass or fail, nothing in between. It passes only when all three measurements hold. A vial that is pure but half empty fails, and so does a full vial of the wrong molecule.

  1. Gate 1

    MS

    Identity confirmed

    The observed mass by mass spectrometry has to match the compound on the label.

  2. Gate 2

    99.5%

    Minimum purity by HPLC

    Main-peak area at the lab’s standard wavelength. 99.0% reads as clean elsewhere; here it fails.

  3. Gate 3

    ±25%

    Fill accuracy

    Measured quantity over the label claim between 0.75 and 1.25.

Where a third party publishes only the outcome (Finnrick does, against purity of at least 99.5% and a quantity within 35% of the label), its outcome is used as published and the row says so.

02

Delivery score

Our estimate of business days from order to a US doorstep, built from each vendor’s published dispatch cutoff and carrier service. One day scores 100 and every further day costs nine points, down to a floor of 20. These are estimates, not measured transit times; a vendor shipping from China by courier is entered at the two to three weeks its own terms describe.

Formula

delivery = clamp(100 − 9 × (days − 1), 20, 100)

Score by estimated business days

0 50 100 1d 3d 5d 7d 10d 15d 1 day, 100 3 days, 82 15 days, 20

03

Price score

Every listed vial is converted to dollars per milligram, so a 5 mg vial and a 10 mg vial of the same compound compare fairly. For each compound we take the median across every listing we track; each vendor’s listing is then a ratio to that median, and the vendor’s score comes from the median of its own ratios, so one dear niche product does not sink a vendor that is cheap on everything else.

At the market median the score is 70. Thirty percent under the median scores 100; the floor is 15. The milligrams are what you receive for the price: kit sellers are entered at their smallest kit, and a standing two-for-one listing counts both vials. Limited-time sales and coupon codes are not applied. Price carries the smallest weight because a cheap vial that fails is not a saving.

Formula

ratio = median over compounds of (vendor $/mg ÷ market median $/mg)
price = clamp(100 − 100 × (ratio − 0.70), 15, 100)

Score by ratio to the market median

0 50 100 0.50x 0.70x 1.00x 1.25x 1.50x 2.00x 0.70x, 100 median, 70 1.55x, 15

04

Purity score

The share of a vendor’s independent samples that passed the standard, weighted by age. A result loses half its weight every 240 days, so a failure last month counts for far more than one from last winter, and a vendor that fixes a problem climbs within a few months.

Thin histories are pulled toward the board-wide pass rate by adding 2 virtual results at that baseline to every vendor. One lucky pass cannot make a perfect score and one failure cannot make a zero. A vendor with no results at all sits at the baseline, currently 72.5. This is the only pillar measured in a lab, which is why it carries the most weight.

Formula

weight = 0.5 ^ (age in days ÷ 240)
purity = 100 × (Σ weight × passed + 2 × baseline) ÷ (Σ weight + 2)

Weight of a result by its age

0 25 50 100 today 240d 480d 720d 240 days, half 480 days, a quarter

05

Trust score

Seven things a buyer can verify before paying, each worth a fixed number of points from a budget of 100. Customer reviews are not among them, because reviews can be bought and lab reports cannot.

  1. 20

    Certificates readable before checkout

    A free account wall still counts; email on request does not.

  2. 20

    Lot on the certificate is the lot on the vial

    Paper that can be matched to the glass in your hand.

  3. 13

    Certificate re-checks on the lab’s own site

    A search code or verify link at the lab, not a PDF alone.

  4. 12

    Testing lab named

    An unnamed "ISO 17025 lab" scores zero here.

  5. 10

    Mass spec identity, not HPLC alone

    HPLC measures purity; it cannot say which molecule it is.

  6. 10

    Trading history

    2 points per year from the registration or launch date, capped.

  7. 15

    No public enforcement actions

    Minus 15 per action on record, floored at zero.

  8. 100

    Total

    Each check is all or nothing except trading history, which accrues two points a year.

06

Overall score

The four pillars combined with fixed weights. Purity carries the most because it is the only pillar measured in a lab; trust is next because it is what lets you check the vendor yourself; delivery and price are what you feel first but matter least if the vial is wrong. On the board, every overall score is drawn as a bar made of these four segments.

Formula

overall = 0.40 × purity + 0.25 × trust + 0.20 × delivery + 0.15 × price

Strong
85 and up
Solid
70 to 84.9
Mixed
55 to 69.9
Weak
under 55

07

Rules of the board

Who is listed

This is a buyer’s guide, so a vendor is listed only when its prices can be read without placing an order. Account walls are fine as long as the store API or the catalog itself can be read; stores that quote only by chat or email, stores whose catalog could not be read at all, raw-material suppliers, and sellers that trade only through Telegram or Discord are left off. The board holds 24 vendors and 12 compounds.

Compound boards

Each compound page lists the vendors with independent results for that compound, ordered by overall score, with the price, delivery estimate, and pass count for that compound alone. One guard: a vendor whose results for the compound mostly fail is listed after every vendor whose results mostly pass, whatever its overall score. Vendors that list the compound without results sit below, unranked.

What is left out

  • Vendor-supplied certificates. Linked for reference, never counted toward purity.
  • Customer reviews and social media. Reviews can be bought; they do not score.
  • Catalog size. Eight well-made products beat eighty inconsistent ones.
  • Limited-time sales and coupon codes. Prices are the regular listed price for the smallest quantity you can buy, counting every vial in a kit or a standing two-for-one listing.
  • Anything whose provenance cannot be shown.

08

Worked example

Solution Peptides is first on the board as of this snapshot. Its score rebuilt from the four pillars on its page, drawn the way the board draws it:

Solution Peptides logo

Solution Peptides

Ranked 1 of 24

95.6

  1. Purity 97.7 × 0.40 39.08
  2. Trust 98 × 0.25 24.50
  3. Delivery 100 × 0.20 20.00
  4. Price 80 × 0.15 12.00
  5. Overall 95.6

09

Data status

Snapshot Oct 28, 2026: 24 vendors, 12 compounds, 626 results. What is measured, what is estimated, and what is a placeholder:

  1. Real

    Vendor names, domains, catalogs, prices, listing links, certificate policy, payment methods, cutoffs, carriers, free-shipping thresholds, published locations, and registration dates, read from each vendor’s own site on 27 and 28 October 2026. Independent results for 18 vendors, which are Finnrick’s published outcomes as of Oct 28, 2026, linked row by row.

  2. Estimated

    Every delivery median. They come from the vendor’s own dispatch cutoff and carrier service, not from measured parcels. Coded product names marked "inferred" in the data were mapped by the usual S / T / R convention rather than confirmed by the vendor.

  3. Placeholder

    The test histories of Solution Peptides and Atomik Labz are sample rows that stand in for our own testing round, and every one of them passes. Their purity scores, and so their positions at the top of the board, rest on those rows. Finnrick holds a real record for Atomik Labz (35 of 44 passed, latest Aug 12, 2026), which is not used here; it holds nothing for Solution Peptides. This line stays until those rows are replaced with verified results.

  4. No results yet

    Peptide Maxxing, Southern Aminos, Ascension Peptides, Biotech Peptides. Finnrick has not covered them, so they sit at the board baseline for purity and are listed for their prices and what they let you verify.