Compass Pathways expects to finish its rolling New Drug Application for COMP360, its synthetic psilocybin for treatment-resistant depression, in the fourth quarter of 2026, and told investors in its August 5 second-quarter results that it expects a commercial launch in the first half of 2027, subject to FDA approval and DEA rescheduling. Helus Pharma expects topline data in the same quarter from APPROACH, the first pivotal Phase 3 study of HLP003, its deuterated psilocin analogue for adjunctive treatment of major depressive disorder.
Neither company makes its own drug or runs the clinics that would administer it. Compass says it has no plans to establish manufacturing facilities and that its strategy depends on third-party treatment sites. Helus says it relies on contract manufacturers. Incannex Healthcare, which is developing the synthetic psilocybin PSX-001 for generalized anxiety disorder, says it does not operate its own manufacturing facilities. The nonprofit Usona Institute, which has completed enrollment in a Phase 3 trial of synthetic psilocybin for major depressive disorder, has not said who would deliver an approved product.
So the practical question about a prescription psychedelic market is who supplies everything around the molecule, and whether any of those suppliers stand to benefit from the category as a whole instead of from one drug. The public record gives a narrower answer than the popular framing suggests. Most of the infrastructure is inherited from esketamine. Most of the disclosed agreements are clinical-supply engagements or research collaborations. None discloses what anyone is paid. And in the filings and releases of these four developers, only one outside provider appears beneath more than one of them.
The clinic base was built for another drug
Johnson & Johnson reported $1,696 million in worldwide Spravato sales for 2025, up 57.4%, and $584 million for the second quarter of 2026, of which $514 million was in the United States. Esketamine nasal spray is administered in certified healthcare settings under a risk evaluation and mitigation strategy, with a two-hour observation period after each dose. Seven years of that requirement produced what psychedelic developers now describe as their delivery network: rooms, monitoring staff, controlled-substance procedures and billing departments that already handle a supervised psychiatric drug.
Compass’s second-quarter release says COMP360 is expected to fit “within current infrastructure of more than 8,000 centers offering multi-hour treatments.” A footnote gives the source: Johnson & Johnson’s public treatment-center locator for Spravato, as of July 31. Helus’s October 2026 corporate presentation reaches “approximately 8,000” by a different route, adding 5,300 Spravato clinics, 750 ketamine-only clinics and 2,300 transcranial magnetic stimulation clinics. Those components sum to 8,350, the presentation cites no source for the Spravato figure, and it makes no adjustment for clinics that offer more than one of the three. Both are company counts built in different ways. We found no count of certified Spravato sites published by FDA or by Johnson & Johnson to check either against, and a listing in a locator says nothing about a site’s spare capacity.
The reimbursement scaffolding is also partly in place and partly promised. The American Medical Association’s Category III CPT codes 0820T to 0822T, for continuous in-person monitoring and intervention during psychedelic medication therapy, took effect on January 1, 2024. They are billed by the hour: one code for the first clinician, and add-on codes for a second clinician and for clinical staff. Category III codes are temporary tracking codes. Medicare’s fee schedule lists these three without a national payment rate, leaving the amount to its regional contractors. At the Deutsche Bank healthcare summit on September 17, according to a published transcript of the session, Compass’s chief patient officer said the codes are live in Category III form and ready to be progressed to Category I, and that the company intends to apply for a J-code after approval. Each of those steps follows approval; none has happened.
The same transcript carries the most useful numbers on clinic economics, all of them management estimates. Compass said it expects a REMS comparable to Spravato’s, with about six hours of monitoring against Spravato’s two. It said sites are currently reimbursed about $250 to $300 for a Spravato visit while blocking a room for at least three hours, and that “probably” 35% to 40% of Spravato is acquired through buy-and-bill, in which the clinic purchases the drug and bills for it.
Put those together and the scarce unit is not the clinic. It is the treatment-room hour. A count of 8,000 sites says little about how many six-hour sessions those sites can absorb alongside an esketamine business that is still growing at 40% a year. That is an inference from the companies’ own figures, not a company statement, and no developer has published a capacity estimate that would test it.
One software company, three developers
Osmind is the clearest case of a provider positioned beneath competing drug developers, and the details show how thin “positioned” still is.
The San Francisco company sells an electronic health record built for interventional psychiatry. Its pricing page lists the record at $249 per clinician per month, with automated Spravato REMS submissions at $4 each. In April 2026 it announced a group-purchasing arrangement with CareNet GPO that begins with Spravato and includes buy-and-bill support. It also licenses de-identified outcomes data: NRx Pharmaceuticals said in January 2026 that it had licensed Osmind data on more than 70,000 patients treated with ketamine or esketamine to support a drug application. Its disclosed products and pricing, in other words, are built around practices that deliver esketamine and ketamine today.
On top of that business it has three developer relationships, and they are not the same deal three times.
Helus, then operating as Cybin, announced a “strategic partnership” on April 21, 2025. The release says the company “will leverage Osmind’s 800-clinic network, point-of-care software, and real-world data,” and describes the work as end-to-end mapping of clinic workflow and patient journey in areas “such as pharmacy, fulfillment, patient access and reimbursement.” Mapping is the operative word. Nothing in the release or in Helus’s subsequent filings says Osmind will operate pharmacy, fulfillment or reimbursement services for HLP003.
Compass announced a collaboration on April 21, 2026, to learn how small and mid-sized independent practices could deliver psychedelic treatments if approved. Its own forward-looking statement calls the arrangement a “research collaboration agreement,” and the direction of support runs the other way: Compass says it “will support Osmind and their network to prepare.”
Tasman Therapeutics announced the broadest of the three on August 12, 2026: a multi-year partnership covering Phase 3 trial support and launch preparation for R-107, an oral extended-release ketamine. It is not a psychedelic, but it is a third developer drawing on the same network.
None of the three announcements discloses payment terms, duration beyond Tasman’s “multi-year,” or exclusivity. Working with several developers establishes that Osmind is not tied to one molecule. It does not establish that Osmind earns anything material from any of them.
The network figure itself deserves care. Osmind’s releases described “over 800 clinics” in spring 2025 and “over 1,000” from April 2026, alternating between “clinics” and “practices” without defining either. Helus’s October 2026 presentation still cites more than 800. The count is self-reported, and no source breaks out how many of those sites administer interventional treatments today.
Manufacturing is contracted, and mostly unnamed
All four developers outsource production. Only one names a manufacturer for late-stage supply.
Helus said on May 15, 2025 that it had engaged Thermo Fisher Scientific for Phase 3 clinical supply and “potential commercial manufacturing” of its lead program, at Thermo Fisher sites in Florence, South Carolina, and Cincinnati, Ohio, the second for capsule production. That is the only agreement in this group with commercial-scale language, and the language is conditional. Helus’s August 2026 management discussion says that “at this time the Company has not entered into commercial supply agreements.” It adds a risk factor worth reading beside the announcement: the company “has not contracted with alternate suppliers for drug substance production.”
Compass, the closest to market, names no manufacturer at all. Its 2025 annual report says it does not own or operate manufacturing facilities and has no plans to establish any, that it relies “on a single supplier for our API,” with back-up suppliers identified, and that COMP360 “is imported in its fully-finished, packaged and labeled dosage form.” The same report notes that, once the drug is approved and rescheduled, importing it for commercial sale requires an importer registration and a permit for each import.
Incannex’s annual report, filed September 25, 2026, says it has engaged Ardena US LLC, formerly Catalent Pharma Solutions LLC, to develop and manufacture oral PSX-001 for “planned clinical trials.” The same filing says the company is adjusting the PSX-001 development plan, including the design of its Phase 2 study. Clerkenwell Health, the UK contract research organization that Incannex said in February 2024 had helped design that study, is not named in the new report. Both are clinical-stage vendors. Nothing in the record makes either a launch partner.
Usona says it completed commercial-scale validation batches in 2025 and describes manufacturing of its active ingredient at two global locations it does not name. It is also a supplier in its own right: its investigational supply program provides its products to outside academic and commercial researchers. As a 501(c)(3) funded by donors, it has disclosed no pricing, distribution or commercialization plan, which leaves open who would own the delivery economics of an approved Usona product.
The pattern extends beyond this group. Definium Therapeutics, developing the lysergide candidate DT120, says in its 2025 annual report that it sources supply “through several CDMOs on a purchase order basis” and has no arrangements in place for commercial or redundant supply. Contract manufacturers are working across the field. Because developers rarely name them, the disclosures reviewed here do not identify one with late-stage work for two competing programs.
What the developers are keeping
The functions developers are building themselves are the ones closest to the prescription: sales, reimbursement support, training and regulatory strategy. Compass has kept a medical liaison team in the field and has told investors it is making sales-force offers contingent on approval and intends to have a field reimbursement team in place at launch. Helus’s October presentation lists a sales force and distribution network and a reimbursement and contracting framework as things to build around HLP003; it appointed a former president of Intra-Cellular Therapies as chief executive on August 3 and is targeting a potential NDA filing in 2028.
Delivery itself stays with third parties, and here Compass has been deliberate. Its annual report lists collaboration agreements with Greenbrook, Hackensack Meridian Health, Reliant Medical Group, Journey Clinical, Mindful Health Solutions, HealthPort and Radial Health, with Osmind added in April. It also states their purpose: “to research and investigate models for the delivery” of COMP360, assuming approval. These are learning arrangements across different care settings. None is disclosed as a commitment to administer the drug, and none comes with site numbers, volumes or exclusivity.
The treatment model determines how much of the inherited base is usable. Compass describes what happens in its trials as standardized monitoring and support, not psychotherapy, and its management has said it expects supervised dosing-site observation of about six hours. That is a staffing and room requirement even without a therapy component, and the label and REMS that will define it do not exist yet. Helus says HLP003’s acute effects last about four to six hours; its presentation charts two clinic visits a year against 34 to 56 for Spravato, and a footnote calls the HLP003 profile illustrative and subject to validation in Phase 3.
Fewer visits per patient is a selling point to payers and patients. For a clinic paid per visit and per monitored hour, it changes the arithmetic in ways the current codes have not priced.
Where exposure sits today
Exposure to a drug’s success and exposure to a category’s growth are different positions, and the record supports three statements about the second.
First, category infrastructure exposure today is esketamine exposure. The clinic operators, the software vendor and the monitoring workflows named in psychedelic developers’ materials are earning from Spravato, ketamine and TMS now. Psychedelics are an option on top of that business, not a revenue line.
Second, the economics are undisclosed everywhere. No partnership reviewed here reports fees, minimums, exclusivity or volume. Preparation for commercialization is not evidence of clinical success, and a named partner is not evidence of revenue.
Third, the visible dependencies run toward concentration, not breadth: a single API supplier at Compass, no alternate drug-substance supplier at Helus, one self-reported clinic network cited by three developers.
None of this ranks the providers. Nothing in the record shows that any of them has secured a position that pays whichever drug succeeds.
The next disclosures will show whether a distinct infrastructure layer is forming. A final label and REMS will fix the staffing and monitoring burden. A Category I code and a J-code will decide whether clinics can be paid for six-hour sessions. The first signed commercial supply agreement will name a manufacturer. And the first research collaboration to convert into contracted treatment capacity will be the first evidence that the clinics beneath these drugs have agreed to more than a conversation.