HOPPR
Infrastructure for building medical imaging AI rather than a clinical application, which makes it structurally different from every other vendor in this lane. Grace is a multimodal foundation model supporting image to image and text to image learning across X-ray, CT, MRI, and echocardiography, trained on over a petabyte of permission based anonymized imaging studies enriched with corresponding reports across 2D, 3D, and longitudinal series. The HOPPR AI Foundry is the surrounding development environment, combining accelerated computing, curated datasets, foundation models, fine tuning tooling, and traceable development workflows inside a stated HIPAA compliant environment, so developers, radiology PACS vendors, and AI companies can build, evaluate, fine tune, validate, and host imaging models without assembling that infrastructure themselves. NVIDIA open models NV-Reason and NV-Generate became available on the Foundry in March 2026, with NV-Reason generating structured analytical reasoning alongside outputs. The MC Chest Radiography Narrative Model, introduced April 2026, is a vision language model translating chest X-rays into structured descriptive text, shipped with training data traceability records and version locking so developers can reproduce results. Forward Deployed Services pairs HOPPR machine learning staff with customer teams for fine tuning. Founded 2019, led by Dr Khan Siddiqui; $34.5 million raised including a $31.5 million Series A in June 2025, with Health2047, the American Medical Association venture studio, among investors.
Capability Axes
The foundation model is the entire product. There is no clinical application to buy, only Grace and the environment for adapting it, which makes this the most AI native position in the index by construction: a customer purchases model weights and the means to fine tune them.
Among the strongest disclosures in the index, and appropriately so given that developers building regulated products downstream inherit whatever the base model does. Training corpus is quantified and characterized: over a petabyte of permission based anonymized imaging studies paired with corresponding reports, spanning 2D, 3D, and longitudinal series across modalities. The narrative model ships with training data traceability records explicitly to support downstream evaluation, and with version locking so teams can reproduce results across development and deployment. Third party models on the platform are named rather than white labeled. Publishing provenance and pinning versions is what a developer needs to defend their own regulatory submission later.
Evidence appropriate to infrastructure is thin here. The narrative model is stated as evaluated against internal benchmarks with no published results, comparison, or external validation, and no downstream clinical application built on Grace was identified with outcome data. Investment from Health2047, the American Medical Association venture studio, and a strategic partnership with RadNet are credibility signals rather than performance evidence. For a base model that others build regulated products on, published benchmark performance is the reasonable expectation and it is absent.
The consent basis for training data is stated rather than implied, described as permission based and anonymized, which is more than most model builders in healthcare disclose, and development runs inside a stated HIPAA compliant environment. Held back from A because the de identification standard, the permission mechanism, and whether customer data used in fine tuning is isolated from the base model were not detailed in retrieved materials. For a platform whose whole business is other people's imaging data, those are the decisive terms.
No clearance is claimed and none would apply, which is correct rather than a gap: HOPPR supplies models and development infrastructure, and regulatory responsibility for any resulting clinical product sits with the developer who builds and submits it. Buyers should understand that clearly, since using a cleared component does not confer clearance and using an uncleared foundation model does not prevent it. The traceability and version locking features exist precisely to support that downstream submission burden.
No public pricing. Contact the vendor. Sold to developers, AI companies, and radiology PACS vendors as platform access, with Forward Deployed Services as a separate professional services engagement pairing HOPPR staff with customer teams. Buyers should establish whether compute, model access, and services are priced separately, since a fine tuning heavy program can shift cost substantially toward the services line.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
Contact the vendor
|
Platform and model access for developers; separate forward deployed services | — | — | Vendor Published |
No rate card published. Sold as platform access to developers, AI companies, and radiology PACS vendors, with Forward Deployed Services offered separately as a professional services engagement pairing HOPPR machine learning staff with customer teams. Establish how compute, foundation model access, and services are priced relative to one another: a fine tuning heavy program can shift the majority of cost into services, and a customer building a regulated product will also carry its own validation and submission expense downstream, which is outside anything HOPPR quotes.