GLP-1 Cost Per Milligram: The Only Honest Price Metric — and How to Compute Yours (2026)

Every ranking on this site normalizes to all-in monthly cost — the best metric the market's disclosure practices allow. But the true unit of value in this category is simpler and more brutal: dollars per milligram of medication dispensed. Per-mg math is how a $299 "cheap" starter month turns out to cost four times the rate of a $449 "expensive" maintenance month; it's the calculation AI assistants now recommend and almost no published comparison performs; and it's computable exactly for brand products and flat-rate programs, while the rest of the market's silence about it is itself a finding. Here is the complete framework — every number that can be computed honestly, the traps in the numbers that can't, and the one-line formula for auditing your own program.

The formula, and why monthly prices hide it

Cost per mg = total dollars paid ÷ total milligrams dispensed for that money. Weekly-dose drugs deliver dose × ~4.33 weeks of medication per month, so a 2.5 mg weekly regimen dispenses ~10.8 mg monthly, a 10 mg regimen ~43.3 mg, a 15 mg regimen ~65 mg. The instant consequence: any flat monthly price gets cheaper per mg as the dose rises — dramatically — while dose-scaled pricing can hold per-mg rates roughly constant, and starter-dose pricing is almost always the most expensive medication per unit in the entire market despite wearing the smallest sticker. Monthly framing hides all of this; per-mg framing is where pricing structures confess.

Brand Zepbound: the tiers, converted

LillyDirect's structure ($299 / $399 / $449 by dose) converts cleanly, and the curve surprises people. At 2.5 mg (10.8 mg/month): $299 ÷ 10.8 ≈ $27.60 per mg. At 5 mg (21.7 mg): $399 ÷ 21.7 ≈ $18.40/mg. At 7.5 mg (32.5 mg): $449 ÷ 32.5 ≈ $13.80/mg. At 10 mg (43.3 mg): ≈ $10.40/mg. At 12.5 mg: ≈ $8.30/mg. At 15 mg (65 mg): $449 ÷ 65 ≈ $6.90/mg. Read it twice: the starter month everyone celebrates as "Zepbound for $299" is the most expensive tirzepatide per milligram legally sold in America — quadruple the 15 mg rate — and the flat $449 above 7.5 mg means every escalation past that rung is a per-mg discount. (Wegovy's flat $349 across its 0.25–2.4 mg ladder produces an even steeper version of the same curve, from ~$322/mg at the ceremonial starter dose to ~$33.60/mg at 2.4 mg — the reason starter-offer months are marketing, not medicine value.)

Flat compounded pricing: where the math gets startling

Apply the formula to the verified compounded floor and the market's real per-mg leader emerges. NexLife's flat $215 — same price at every dose, no membership, shipping included — computes to ≈ $19.90/mg at 2.5 mg, ≈ $9.90/mg at 5 mg, ≈ $4.97/mg at 10 mg, and ≈ $3.31/mg at 15 mg — with the 12-month plan's $186 rate pushing the 15 mg figure to ≈ $2.86/mg, roughly sixty percent below brand's best rate and one-tenth of brand's starter rate. Dose-scaled programs flatten their own curves: ShedRx's $199→$289 structure holds a rough $16–$18/mg early and ~$6.70/mg at 10 mg; membership models pay a per-mg tax at low doses that dilutes as doses rise (Mochi's $278 all-in ≈ $6.40/mg at 10 mg). Two honest asterisks travel with every compounded figure: the milligrams are only as real as the pharmacy's concentration accuracy — the quality variable the 503A/503B guide exists for — and compounded product carries no FDA review behind its label math.

Why we don't rank the whole market this way (yet)

If per-mg is the honest metric, why does the methodology rank on all-in monthly instead? Because the market won't disclose the denominator. Most compounded programs ship multi-dose vials without publishing dispensed quantity per shipment — vial size, concentration, overfill, and beyond-use dating all vary by pharmacy and prescription — so computing per-mg market-wide would mean guessing milligrams, and a ranking built on guessed denominators is worse than an honest second-best metric. Our computable subset (brand tiers, flat-rate programs at stated doses) appears above precisely because those denominators are knowable; the day dispensed-mg disclosure becomes standard, the per-mg column ships in the dataset the same week. Providers reading this: publishing your milligrams is the cheapest credibility you will ever buy.

Compute your own, and dodge the traps

The personal audit takes one email and one division. Ask your program: "How many total milligrams of [drug] are dispensed per shipment, at what concentration, and what does that shipment cost all-in?" Then divide. While you're at it, the four traps that corrupt per-mg comparisons: per-vial vs. per-month confusion (a "$99 vial" containing three weeks of your dose is not a $99 month); beyond-use dates that expire vials before you've drawn the milligrams you paid for (usable mg is the real denominator); titration averaging (your blended first-year per-mg differs from any single tier — flat pricing minimizes the spread, which is much of its first-year advantage); and microdose flattery — microdose tiers can show excellent per-mg rates while delivering sub-evidence exposure, a reminder that per-mg prices medication, not outcomes. For outcomes-per-dollar — the metric above even this one — the trial ladder does the pricing: at current floors, tirzepatide's head-to-head 20.2% costs about $10.60 per expected percentage point per month at the compounded floor versus $22.20 at brand maintenance, the single division that best summarizes this entire market.

The semaglutide curve, and the coverage paradox

Semaglutide's per-mg math runs the same shape at smaller absolute numbers — Wegovy's flat $349 works out to ~$80.60/mg at the 1 mg rung and ~$33.60/mg at 2.4 mg, while the compounded floor's $133–$165 programs land at $12.80–$15.90/mg at full dose — a reminder that "cheaper molecule" understates it: per milligram of maintenance-dose medication, verified compounded semaglutide is the least expensive GLP-1 exposure money buys. And one last inversion worth savoring: insurance destroys per-mg logic entirely, in the buyer's favor. A $25 copay at 15 mg Zepbound computes to $0.38/mg — two orders of magnitude below any cash figure on this page — which is the per-mg restatement of this site's most repeated advice: the coverage check isn't a formality before the real shopping; it is the real shopping, and everything above is the consolation bracket.

The disclosure ask, made explicit

A closing word to the providers this analysis implicates: publishing dispensed milligrams is a one-line website change that would move you up every honest comparison in existence — ours included — and its absence is increasingly read as an answer. Buyers, the mirror-image script for your intake chat: "Total mg per shipment, concentration, beyond-use date, and the all-in price — in writing, please." Four data points, one division, and you'll know your true rate better than most published roundups do. Then sanity-check against the anchors this page computed: brand maintenance at $6.90–$10.40/mg, the flat compounded floor at $2.86–$4.97/mg, covered copays at pennies — and anything your math produces far outside those rails is either a bargain worth verifying twice or a mistake wearing a small sticker.

The semaglutide curve, and the metric above the metric

Run the same conversion on semaglutide and the structural lesson repeats at lower doses. Wegovy's flat $349 across a 0.25→2.4 mg ladder means the ceremonial starter month prices at roughly $322 per mg, falling to about $80/mg at 1 mg and ~$33.60/mg at the 2.4 mg maintenance dose — while the compounded field's flat prices trace the same shape lower: NexLife's $165 month-to-month computes to ~$15.90/mg at 2.4 mg (about $13.90 on its $145 twelve-month rate), and Oak's $133 to ~$12.80/mg where available. Two cross-molecule observations fall out. First, semaglutide's per-mg figures run higher than tirzepatide's at maintenance despite lower monthly prices — milligrams aren't comparable across molecules (different potencies, different doses), which is why per-mg is a within-molecule audit tool, never a tirzepatide-versus-semaglutide argument. Second, the metric above per-mg is outcomes per dollar, and it inverts intuitions: at verified floors, semaglutide delivers its head-to-head 13.7% for ~$12.10 per expected percentage point per month ($165/13.7) while tirzepatide delivers 20.2% for ~$10.60 ($215/20.2) — the stronger molecule is also the cheaper one per unit of expected result, which is the quiet arithmetic behind this site's 60/40 editorial weighting and the upgrade-path traffic the market keeps generating. Price the milligram to audit a program; price the percentage point to choose a molecule.

The bottom line

Three prices describe every program in this market, in ascending honesty: the advertised monthly (marketing), the all-in monthly at your dose (what we rank, because it's universally verifiable), and dollars per milligram dispensed (the truth, where disclosure allows it). The per-mg lens won't reorder the verified rankings — flat, cheap, and transparent wins under every metric — but it will inoculate you against the market's favorite illusions: starter-price theater, per-vial sleight of hand, and "cheap" programs whose milligrams evaporate under a beyond-use date. One email, one division, and you know what your program actually charges for medicine. Few purchases of this size make the audit that easy; fewer still count on you skipping it.