The Patent Cliff and the First Generic Peptide Blockbuster

On 21 March 2026, the day after the compound patent on semaglutide expired in India, several of the country's largest drugmakers put their own versions of it on sale. Within days the listed prices ran from a few hundred rupees a month to a little over four thousand, and the Indian press was describing a price war 4. A molecule that had carried a small European country's growth figures had become, in the world's most populous nation, a generic.

It was not the first place the molecule went off-patent, and it would not be the last. The expiry reached China and Brazil on the same date, some time after the patent had already lapsed in Canada over an unpaid fee, and years before protection is due to end in the United States and most of Europe. This piece is about that staggered cliff: how patents on a peptide medicine differ from those on an ordinary pill, why the protection ends at different times in different places, and what the first generic peptide blockbuster is doing to price and to manufacturing. How the industry that will make the generics came to sit where it does is told separately on this site.

What a patent on a peptide covers

A modern medicine is protected by a thicket rather than a wall. The foundational patent claims the molecule itself — its sequence, its modifications, the chemical idea that makes it new. Around it grow later patents on formulations, on methods of manufacture, on particular uses and, for an injected medicine, on the device that delivers it. The thicket matters in litigation, but in most markets it is the compound patent whose end lets competitors in, because once the molecule is free a well-resourced generic maker can usually design around the rest.

Peptide medicines add a complication that small-molecule drugs lack. A generic tablet is a chemical and a filler. An injectable peptide is a molecule, a sterile solution and a device — a pen filled in an aseptic facility — and each of those is a barrier to entry even when no patent protects it. The expiry of the compound patent removes the legal barrier. It does not remove the industrial ones, which is why the competition that follows a peptide patent cliff is likely to arrive more slowly, and from fewer companies, than the classic small-molecule cliff, where dozens of firms can launch within weeks.

Patent terms in most systems run twenty years from filing, with extensions to compensate for time lost in regulatory review: patent-term extensions in the United States, supplementary protection certificates in the European Union. Those extensions are why semaglutide's protection in its two richest markets is expected to run into the early 2030s 8, while countries that grant no comparable extension — or where the patent was not kept up — saw it end in 2026.

Drug or biologic: a line at forty amino acids

Before a generic can be approved, somebody has to decide what kind of product it is copying. In the United States that question has a numerical answer. A rule that took effect in March 2020 defined a protein, for the purposes of biologics regulation, as an amino-acid polymer with a specific defined sequence greater than 40 amino acids in size 7. Anything shorter is a peptide, and is regulated as a drug.

The consequence is large. A copy of a protein medicine must go through the biosimilar pathway, which typically demands extensive comparative studies. A copy of a peptide medicine can use the abbreviated generic pathway, which asks the applicant to show that the active ingredient is the same and that the product's quality is comparable, rather than to repeat the clinical programme. Semaglutide, at 31 residues, sits on the drug side of the line, as do liraglutide, exenatide and tirzepatide.

The American regulator rehearsed the pathway on the older members of the family. In November 2024 it approved the first generic of exenatide and, the following month, the first generic of the once-daily liraglutide injection, citing in part the shortages then affecting the class 5. In July 2026 it went further. It withdrew a 2021 guidance that had been confined to synthetic copies of five peptides originally made by recombinant DNA technology, and published 17 revised draft product-specific guidances for peptide products, including semaglutide and tirzepatide. The new guidance covers recombinant, synthetic and semi-synthetic routes, and addresses impurity thresholds and testing for innate immune responses 6.

One molecule, no common timetable

Canada went first among wealthy countries, and by accident. Its patent system requires annual maintenance fees; the company let a fee on its semaglutide patent go unpaid, the grace period passed, and the patent lapsed — and, under Canadian rules, a lapsed patent cannot be revived 3. Generic makers filed as the nominal expiry approached. On 28 April 2026 Health Canada authorised the first generic semaglutide injection, for type 2 diabetes, and noted that Canada was the first G7 country to do so 1. On 29 June it authorised the first generic for chronic weight management, with six further submissions still under review 2.

India's expiry on 20 March produced the fastest response anywhere, because India has one of the world's largest generic industries and its companies had been preparing for years. Launches came on the first day from several of the biggest domestic firms, and prices began falling among the launchers themselves almost at once 4.

Brazil's patent expired the same day. There the first approval took two months: on 26 May 2026 the national regulator, Anvisa, registered a semaglutide from the domestic manufacturer EMS, indicated for type 2 diabetes. Local press reported it selling at roughly half the price of the branded product, with further approvals following through the year 10.

China's patent also ended on 20 March, but approval did not follow on the same schedule. Domestic applications were pending, some of them filed well before the expiry, yet by May the originator said it expected generic competition in China to be delayed until the following year 11. The United States and most of the European Union remain on the far side of the cliff, with protection expected to last into the early 2030s 8.

Yeast, resin and solvent

The generic wave is also a manufacturing story. The originator makes the backbone of semaglutide by fermentation in engineered yeast, then attaches the fatty-acid side chain and other modifications chemically. Most generic makers are not replicating that process. They are building the whole chain by chemical synthesis — one residue at a time on a solid support, or by making fragments in solution and joining them, an approach that can raise yield and cut solvent use 8.

That choice has two consequences. The first is regulatory. A synthetic copy of a molecule the originator made partly biologically will not carry the same impurities, and showing that its impurities are no more numerous, and no more likely to provoke an immune response, is the heart of what the revised American guidance asks for 6. The question of sameness, which for a small-molecule generic is largely settled by showing the same structure, becomes for a peptide a question about everything that travels with the structure.

The second is industrial. Demand for generic semaglutide across several of the world's most populous countries amounts to a very large new order for synthetic-peptide capacity, and for the solvent, the purification columns and the sterile fill-finish lines that come with it. How that capacity came to be concentrated in a small number of chemical clusters, and how the shortage years exposed the fill-finish layer, are told elsewhere on this site. The 2026 expiries are the first time that industrial base has been asked to supply a blockbuster at generic prices.

What a generic does to a price

The effect on price is the part everyone was waiting for, and it is visible already. Announcing its first approval, Canada's health department noted that generic medicines in Canada are often 45 to 90 per cent cheaper than the brand-name versions 1. Brazilian pharmacies priced the first local version at about half the branded product 10, and the Indian launches opened a price war among the launchers themselves within days 4.

The originator's accounts registered the change. Its outlook for 2026, published in February, listed the loss of exclusivity for the semaglutide molecule in certain international markets among the reasons it expected sales to fall 9. For a company whose growth had moved a national economy, that single line marked the start of a different kind of business, in which its best-known molecule is still a blockbuster in some countries and already a commodity in others.

What happens next is less certain than the small-molecule precedent suggests. Peptide generics cost more to make than generic tablets, need sterile capacity that is slow to build, and face regulators still settling how to compare a synthetic copy with a partly biological original. Prices will fall. How far, and how fast, depends on how many companies can clear the industrial barriers that remain once the legal one is gone — and on whether the capacity built for India, Brazil and Canada is ready when the cliff reaches the United States and Europe.

Where the story stops

The patent cliff answers the commercial question of when. It leaves open the analytical one: how a regulator decides that a peptide made one way is the same medicine as a peptide made another. That turns on impurity profiling at very low levels, on characterising closely related sequences that differ by a single residue or a single modification, and on assays for immune responses that no small-molecule generic has ever had to pass. It is where a technical article would begin.