Superluminal Medicines

A judgement call between this grade and the top one, and it lands here for a reason worth stating. The company consistently presents three co equal pillars, structure based drug discovery, protein dynamics, and machine learning, and two of those three are physics disciplines that long predate modern machine learning.

Structure based design and molecular dynamics were producing drugs before generative models existed, and part of the structural work here runs on licensed third party software rather than on the company's own models. Apply the test this axis uses and the honest answer is that removing the machine learning would leave a computational chemistry company, which is a real business.

Against that, the company describes itself as a generative biology and chemistry company in every announcement, and its stated differentiator is the speed from prediction to validated structure to biological hits, which is a claim about models. Genuinely central, not solely so.

The architecture is right and the evidence for it is not published. The platform is described as predict, design and test, with experimental validation named as a component rather than implied, so a physical checkpoint exists between what the models propose and what the company claims. That is the same structure that earned a higher grade at two neighbours in this lane.

The difference is that those vendors published numbers from inside the loop, showing how often designs fail as well as how often they succeed, and Superluminal publishes none. No hit rate, no pass rate, no round by round data and no statement of who selects which designs are synthesised or what happens when a prediction and an assay disagree.

Components are named with real specificity and none of them is documented. The platform is called Hyperloop and the parts are identified individually: structure based drug discovery, protein dynamics modelling, generative chemistry and machine learning, in silico pharmacokinetic and toxicology prediction, and experimental validation, organised as predict, design and test.

Naming five distinct capabilities is meaningfully more than the platform label most of this lane offers, and it lets a reader reason about what the system does. Beyond that nothing: no architecture, no model class, no training data, no version, no benchmark and no publication of any kind. The claim to an industry leading in silico prediction capability for pharmacokinetics and toxicology is self assessed against no named comparator and no published measurement.

Among the best in this lane, and earned on named dependencies rather than on policy. Two external systems are identified with their specific functions, Schrodinger LiveDesign for digital molecule design and shared analysis of experimental and virtual data, and CDD Vault for protocol definition and assay data organisation.

Naming the actual tools a discovery stack runs on is rare across this entire index, and it lets a reader see that part of the structural and design capability is licensed rather than built. What remains undisclosed is the proprietary half: no training data provenance for the generative and predictive models, no compute or hosting provider, no foundation model base, and no open source component inventory. Nvidia's venture arm is an investor across both rounds, which implies a compute relationship that is nowhere stated as one.

The commercial validation is among the strongest in this lane and it carries a conflict that has to sit alongside it. Eli Lilly invested in the Series A in September 2024 and then, twelve months later, signed a collaboration worth up to $1.3 billion with a further equity investment and exclusive development and commercialisation rights. A pharmaceutical company committing twice, with its own money at risk both times, is harder to manufacture than any case study.

But Lilly is investor, partner and landlord, since the company operates from Lilly Gateway Labs, so this is not independent validation and should not be read as such. Beyond the deal there is little to check: two dedicated passes on 29 August 2026 located no publication, preprint, benchmark or disclosed compound, no target or indication has been named, and the wholly owned candidate said to be entering human trials during 2026 has no located trial registration or dosing announcement.

No protected health information is in scope. The separation question that applies here is sharper than the usual version in this lane, because of how close the parties sit. Superluminal runs a wholly owned pipeline in the same target family and the same therapeutic area, cardiometabolic disease, that its Lilly collaboration covers, and it does that work inside Lilly's own incubator building while Lilly holds equity in it.

Nothing published states how the wholly owned programmes are walled from the partnered ones, what the platform retains from collaboration work, or how target selection avoids collision between a partner's undisclosed targets and the company's own. Physical colocation with the partner makes the question more pressing rather than less. Recorded as a located absence: nothing indicates a problem, and nothing addresses it either.

Graded neutrally because the obligation has not yet arisen. The platform operates on receptor structures, molecular dynamics and assay measurements, with no protected health information in scope and no covered entity relationship.

Recorded in advance so a later reader does not have to rediscover it: the company stated an intention to bring a wholly owned candidate into human trials during 2026, and a sponsor running its own trials acquires patient data obligations that nothing published anticipates. If a trial opens, this axis should be regraded on the arrangements then in force rather than left at a neutral mark earned during discovery.

Two dedicated passes on 29 August 2026 across a targeted search and a direct review of the navigation and footer found no security page, no trust centre, no certification and no attestation. This vendor goes further than most of the lane in one respect: the site carries no legal documents at all, with no privacy policy, no terms of use and no cookie policy anywhere in the footer or navigation, so there is not even the standard disclaimer language that usually constitutes a company's entire published security posture. It holds undisclosed target information for one of the largest pharmaceutical companies in the world and publishes nothing whatever about how it is protected.

Nothing cleared and nothing adverse. All programmes are preclinical on the public record. A wholly owned candidate outside the Lilly collaboration was expected to enter human trials during 2026, and as of 29 August 2026 two dedicated passes located no investigational application, trial registration, first dosing announcement or designation.

One data quality warning belongs here rather than anywhere else: at least one commercial drug pipeline database attributes an unrelated clinical stage compound and its trial results to this company. Nothing the company has disclosed supports that attribution, and a reader checking regulatory status through aggregators should verify it against primary sources.

Nothing exists on any published surface as of 29 August 2026. The site runs to about, science, pipeline, careers and press sections and none addresses model governance, evaluation practice, failure modes or responsible use, and there is no model card or validation methodology. Unlike several neighbours in this lane, no body of published method work stands in the gap, because this company has published nothing at all.

The relevant failure mode is concrete and unaddressed: a platform built to predict which structural change produces a therapeutic effect will have target classes and conformational states where it performs poorly, and no statement exists of where the approach degrades or how the company would know.

There is no document to grade. No terms of service, no privacy policy and no legal notice of any kind appears on any page of the site as of 29 August 2026; the footer carries navigation links, an address, a telephone number and a copyright line. Consequently there is no warranty position, no limitation of liability, no indemnity and no published allocation of rights over designed compounds.

The single visible fragment of rights allocation comes from a press announcement rather than a legal document: Lilly holds exclusive development and commercialisation rights to compounds arising from the collaboration, with royalties flowing back. That describes how one negotiation was settled and tells a reader nothing about how another would be.

No electronic health record surface exists and none would be appropriate for a preclinical discovery company, so this is graded neutrally under the convention for this lane. Recorded in its place, and it is the most concrete answer to this question anywhere in this cluster: the company names the third party software its scientists actually work in, using Schrodinger LiveDesign for molecular design, data access and collaboration, and CDD Vault for protocol definition and assay data management.

Those are checkable dependencies with named functions, and Schrodinger is separately indexed here. No application programming interface or product surface is offered to customers, because nothing here is sold as software.

Nothing is published about where partner or programme data is held, under what separation, with what retention or with which subprocessors, and no software is deployed to anyone. Graded neutrally under the convention for this lane, now the tenth consecutive record where that convention rather than the evidence sets the grade.

The one deployment fact that is disclosed is physical and consequential: the company's registered address is inside Lilly Gateway Labs, the incubator operated by its investor and largest commercial partner. Sharing a building with a counterparty is a data adjacency question in its own right, and nothing published addresses how systems, records or access are separated.

Better than most of this lane on structure and no better on amounts. The Lilly agreement is broken into its components rather than quoted as a single headline: upfront and near term payments, an equity investment, development and commercial milestones, and tiered royalties on net sales, with the rights allocation stated plainly, Lilly taking exclusive development and commercialisation of collaboration compounds.

Naming what a biobucks figure is made of is a real disclosure and most vendors here do not do it. Both funding rounds are dated, sized and attributed to named leads and full syndicates, and the programme count is stated at six. What holds the grade down is that no component of the $1.3 billion is separately sized, so the headline remains a ceiling that may never be paid, and no target, indication or compound has been disclosed on either the partnered or the wholly owned side.

Narrow by design and explicit about the boundary. The work is small molecules against membrane proteins, starting with G protein coupled receptors and with a stated intention to extend to ion channels and transporters. Therapeutic focus is cardiometabolic disease and obesity on the partnered side, with six small molecule programmes reported overall. Nothing here reaches biologics, other target classes, translational work or clinical development, and the company does not claim otherwise. The choice is defensible on its own terms, since the undrugged fraction of this receptor family is large enough to occupy a company indefinitely, but it is a narrow surface for a buyer.