Three of them, and none is for sale. Aleniglipron, elecoglipron and TERN-601 have all published human data in the past year. The only oral GLP-1 a reader can actually buy is semaglutide in tablet form, whose economics are set out in what the tablet costs against the injection. The rest of this page is what each of the three has shown, and what is still missing from every one of them.
Aleniglipron, the newest readout
ACCESS was a phase 2b trial that randomized 230 adults with obesity or overweight to once-daily aleniglipron at 45, 90 or 120 mg, or to placebo [1]. Average body mass index at entry was 39.5, the dose was escalated every four weeks, and the trial ran 36 weeks before the open-label extension began.
Placebo-adjusted weight change was 8.2% at 45 mg, 9.8% at 90 mg and 11.3% at 120 mg, each at p<0.0001. The phrasing matters more than the figures do. Placebo-adjusted means the placebo arm’s own change has already been subtracted, so the number a marketing page would print is the arm change, which is larger and which this abstract never gives.
Treatment-related discontinuations ran 10.4% across the drug arms. Gastrointestinal events were mild to moderate and became less frequent over time, and the authors report little or no recurrence of vomiting when the drug was restarted after a permitted interruption. No drug-induced liver injury occurred, which is not a throwaway detail in this class.
Elecoglipron, and why it is a blood sugar result
SOLSTICE randomized 404 adults with type 2 diabetes across eight groups, testing oral elecoglipron, placebo, or open-label oral semaglutide over 26 weeks [2]. HbA1c fell between 0.91 points at 5 mg and 1.88 points at 75 mg with two-weekly escalation, against 0.15 on placebo, and the fuller dose-by-dose reading sits in the eight-arm tablet trial.
Two cautions travel with those numbers. The primary endpoint was blood sugar and no weight outcome appears in the abstract at all, so anyone reading this as evidence about a future weight-loss tablet is extrapolating from something nobody measured. The semaglutide arm was also open-label while every elecoglipron arm was masked, which rules out reading the trial as a head-to-head.
The tolerability pattern is the part that transfers. Adverse events ran 63% at the lowest dose against 87% at the highest with the fastest escalation, and the 63% figure is identical to placebo. Side effects tracked the dose and the speed of titration rather than the drug itself, which makes the schedule a lever rather than a formality.
TERN-601, the one that stopped
TERN-601 was tested in a 28-day phase 1 and a 12-week phase 2, both randomized, double-blind and placebo-controlled [3]. It produced dose-dependent weight loss, significant against placebo at every dose in phase 1 and at 500 mg or more in phase 2, with the appetite and gastric-emptying signature of the class.
Three participants in the phase 2 study then had transaminase elevations consistent with potential drug-induced liver injury, and development was discontinued. Three is not a small number at that stage. A phase 3 program enrolls thousands and a marketed drug reaches millions, so a signal visible in a few dozen people would be visible in a great many more. The full account sits in the oral trial that was halted.
The only pill anybody can buy
Oral semaglutide is the tablet that exists, and a systematic review of 30 studies gathered what is known about it [4]. The figures everybody quotes are −15.1% at 68 weeks in OASIS 1 for the tablet and −14.9% in STEP 1 for the injection. Those come from two separate trials, with different participants in different years. The review states plainly that head-to-head trials are lacking.
One number explains the price. Oral semaglutide absorbs at roughly 1% against roughly 89% for the injection, so the tablet is dosed at 50 mg daily where the injection is 2.4 mg weekly. That is about twenty times the active ingredient per dose, and it is why a format with no needle and no cold chain does not come out cheaper, as the dose comparison works through.
What a buyer should do with a pipeline
Treat it as a probability distribution rather than a schedule. TERN-601 worked and still stopped, which is the ordinary failure rate of drug development rather than an unlucky exception, and the same uncertainty attaches to the strongest candidate still in trials.
That is not an argument for buying now either. It is an argument against signing anything long on the assumption that a cheaper tablet arrives on a particular date, because no date exists. What the market charges today, for the formats that are actually on sale, is counted in what a GLP-1 costs per month.