FGP — independent test log
Every third-party result a GLP Research Hub member has paid for and published on FGP. Vendor-supplied COAs are not on this page — only tests where a member sent a vial to Janoshik, Medutest, PeptideMeter or VendorInvestigate themselves.
| Date | Service | Compound | Batch | Claimed | Result |
|---|---|---|---|---|---|
| 2026-06-10 | Janoshik | Cagrilintide 5mg | FGP-2606 | 98.0% | 98.3% |
| 2026-01-15 | PeptideMeter | Semaglutide 10mg | FGP-2601 | 98.5% | 98.1% |
| 2025-08-04 | Medutest | Tirzepatide 10mg | FGP-2508 | 98.0% | 98.4% |
| 2025-02-27 | Janoshik | Retatrutide 5mg | FGP-2502 | 98.0% | 97.6% |
How to read this table
Three columns matter and they are easy to conflate. Claimed is what the vendor documentation said. Result is what an independent service measured on a vial a member actually received. The gap between them is the only number on this page with any predictive value — the absolute purity figure tells you much less than whether the vendor's own claim survives contact with a lab.
A vendor claiming 98% and delivering 98% is more trustworthy than one claiming 99.5% and delivering 98.6%, even though the second vial is nominally better. Consistency beats peak performance when you cannot test every order.
Also note what this table is not: it is not a survey. These are retail-quantity purchases by individual members, self-selected, self-funded and self-reported. FGP has 4 entries here; the vendors with more entries look more variable partly because more sampling reveals more variance.
Services used
All four independent services appear across the vendor directory: Janoshik (HPLC and mass spec, the most-used by members), Medutest, PeptideMeter (which also publishes vendor-level trend data) and VendorInvestigate. Where two services tested the same batch we list both rows rather than averaging them, because the disagreement is the interesting part.