ACHIEVE-3 published in The Lancet with a 73.6% greater relative weight loss

Eli Lilly published the complete results of ACHIEVE-3 in The Lancet on February 26, 2026, the first head-to-head Phase 3 trial of two oral GLP-1 drugs in adults with type 2 diabetes. Orforglipron beat oral semaglutide (sold as Rybelsus) on blood sugar and weight, but more people stopped orforglipron because of side effects. ^[1][4]

The 52-week, open-label trial randomized 1,698 adults with type 2 diabetes that was not well controlled on metformin. Participants were assigned in equal numbers to orforglipron 12 mg or 36 mg, or oral semaglutide 7 mg or 14 mg, at sites in the U.S., Argentina, China, Japan, Mexico and Puerto Rico. Average starting A1C was 8.3% and average starting weight was 97.0 kg (213.9 lbs). ^[1]

What the numbers showed

On the primary endpoint, orforglipron 36 mg lowered A1C by 2.2 percentage points at week 52 versus 1.4 with oral semaglutide 14 mg. The lower doses followed the same pattern: 1.9 with orforglipron 12 mg versus 1.1 with oral semaglutide 7 mg. ^[1]

Weight loss, a secondary endpoint, was 9.2% (8.9 kg; 19.7 lbs) with orforglipron 36 mg compared with 5.3% (5.0 kg; 11.0 lbs) on oral semaglutide 14 mg — what Lilly describes as 73.6% greater relative weight loss. Orforglipron 12 mg produced 6.7% (14.6 lbs) versus 3.7% (7.9 lbs) with oral semaglutide 7 mg. Using the more conservative treatment-regimen estimand, which counts everyone regardless of whether they stayed on the drug, the 36 mg figure was 8.2% (17.8 lbs) versus 5.3% (11.5 lbs). ^[1]

More participants on orforglipron reached tight glucose targets. An A1C under 5.7% — the non-diabetes range — was reached by 37.1% on orforglipron 36 mg versus 12.5% on oral semaglutide 14 mg. For A1C under 7%, the figures were 85.4% versus 66.1%. Lilly notes the under-5.7% endpoint and the 12 mg weight comparison were not controlled for family-wise type 1 error, meaning those results carry more statistical uncertainty. ^[1]

The tolerability trade-off

The side-effect picture ran the other way. Discontinuation because of adverse events was 8.7% on orforglipron 12 mg and 9.7% on 36 mg, versus 4.5% and 4.9% on oral semaglutide 7 mg and 14 mg — roughly double. Reuters reported that about 58% of patients on 36 mg orforglipron had mild-to-moderate side effects including nausea, diarrhea and vomiting, compared with 45% on 14 mg oral semaglutide, and that orforglipron users had an average increase in pulse rate. ^[1][4]

The most common adverse events in both groups were nausea, diarrhea, vomiting, dyspepsia and decreased appetite. ^[1] Lilly's Kenneth Custer told Reuters the company views this as "a trade-off that patients will be very happy to make because they are getting better glycemic control and better weight loss as well as no restriction on dosing with food and water." ^[4] Patients and clinicians may weigh that differently.

Orforglipron is a small-molecule pill that can be taken any time of day without food or water restrictions. ^[1] Oral semaglutide requires specific timing around food and water. Both arms used a stepwise dose escalation every four weeks, and participants who could not tolerate a dose were allowed one step-down during the study. ^[1]

Why it matters for patients

For people with type 2 diabetes who prefer a pill over an injection, this is the first direct comparison of two oral GLP-1 options in the same trial, with the same protocol and the same population. Until now, comparisons across separate studies were guesswork. ^[1]

The practical picture: on these doses, orforglipron delivered roughly a percentage point more A1C reduction and about 9 extra pounds of weight loss over a year, while roughly twice as many people quit because of side effects. ^[1] The trial was open-label, meaning participants knew which drug they were taking, which can influence reported symptoms. It also did not test higher weight-management doses of semaglutide such as Wegovy, so it says nothing about how orforglipron compares to injectable semaglutide or tirzepatide. Long-term cardiovascular outcomes were not measured here.

What happens next

Lilly says it has submitted orforglipron to regulators in more than 40 countries, with potential U.S. FDA action for obesity in the second quarter of 2026 and a U.S. submission for type 2 diabetes planned later in 2026. ^[1] Detailed results from the three remaining ACHIEVE registration trials are expected later this year. ^[1] Pricing and insurance coverage are not yet known.