$150M BARDA contract. DeepView FDA De Novo application submitted; determination pending. Revenue of $19.7M in FY2025 down from $29.6M in FY2024 due to contract milestone timing, not commercial failure. $15.4M cash at year-end.
Spectral AI's revenue model is almost entirely funded by BARDA; the Biomedical Advanced Research and Development Authority; through a contract worth up to $150M, with an additional $31.7M award in 2025. FY2025 R&D revenue came in at $19.7M, down from $29.6M in FY2024. Management's FY2026 revenue guidance is $18.5M. The year-over-year decline is not a commercial failure; it is contract milestone timing. BARDA contracts pay on deliverables, not calendar time.
The DeepView FDA De Novo application is the pivotal event. De Novo is the FDA authorization pathway for novel medical devices without a predicate; more rigorous than 510(k), less extensive than full PMA. The application has been submitted and is awaiting determination. UKCA marking for the UK market is already in place. The FDA determination timeline is not publicly specified but a positive outcome expands the commercial pathway dramatically.
The Human Translation: The Government-Funded Diagnostic Device. BARDA is the U.S. government's medical countermeasures agency. When BARDA signs a $150M+ contract with a medical device company, it has done independent technical diligence on the underlying science. The DeepView system has already passed BARDA's evaluation. The FDA De Novo application is the next door to open; and BARDA's prior validation makes the application stronger.
Q4 2025 gross margin (the share of revenue left after direct costs) was 39.8 percent, down from 44.0 percent in Q4 2024. The margin compression reflects the revenue mix within the contract milestones for the quarter. Cash at December 31, 2025 was $15.4M, up from $10.5M at September 30, 2025 and $3.7M a year ago; the BARDA funding is building the cash position.
The commercial hospital market is the long-term opportunity. The military pathway is the near-term proof point. A multi-site military deployment would simultaneously validate the technology through the most demanding institutional customer in the world and generate incremental revenue above the BARDA contract baseline.
The BARDA relationship is the foundational analytical fact for Spectral AI. BARDA; the Biomedical Advanced Research and Development Authority, is not a typical government customer. It is the U.S. government medical countermeasures agency responsible for ensuring that the country has the diagnostics, therapeutics, and preparedness infrastructure needed for public health emergencies, military medical operations, and mass casualty scenarios. When BARDA signs a $150M-plus contract with a medical device company, it has performed independent technical diligence on the underlying science, the manufacturing capability, the regulatory pathway, and the operational readiness of the company. BARDA contracts are not awarded based on press releases or investor presentations. They are awarded based on demonstrated technical capability evaluated by government scientists and medical professionals. The contract itself is external validation of the DeepView technology at a level that most medical device companies never achieve.
The DeepView system uses multispectral imaging and artificial intelligence to assess burn wound depth in real time; a diagnostic challenge that currently has no reliable non-invasive solution. The current standard of care for burn depth assessment is clinical observation, which requires waiting 48 to 72 hours for the wound to declare itself; essentially waiting for the tissue to show signs of healing or necrosis before determining the severity. During that waiting period, treatment decisions are delayed, patients may receive inappropriate interventions, and resource allocation in mass casualty scenarios is suboptimal. DeepView provides an assessment within minutes of imaging, allowing clinicians to make treatment decisions immediately. In a military context; where burn injuries are common and evacuation resources are limited; the ability to triage burn patients accurately and immediately has direct implications for survival and resource allocation.
The FDA De Novo authorization pathway is the regulatory milestone that determines Spectral AI commercial timeline. De Novo is the FDA pathway for novel medical devices that have no predicate; there is no existing cleared device that does what DeepView does, so the standard 510(k) pathway, which requires demonstrating substantial equivalence to a predicate device, is not available. De Novo requires the company to demonstrate that the device is safe and effective for its intended use through clinical data, bench testing, and manufacturing quality documentation. The application has been submitted and is awaiting determination. UKCA marking for the UK market is already in place, which means a parallel regulatory authority has already evaluated the device and found it acceptable for clinical use. The FDA determination is not guaranteed, but the BARDA contract and the UKCA marking provide supporting evidence of the technology validity.
The revenue model during the BARDA contract period is fundamentally different from a traditional commercial medical device revenue model. Spectral AI recognizes revenue based on achievement of contract milestones; specific technical deliverables, clinical study completions, manufacturing readiness assessments, and other defined objectives within the BARDA scope of work. This means revenue fluctuates based on milestone timing rather than unit sales. FY2025 revenue of $19.7M versus FY2024 revenue of $29.6M reflects the cadence of milestone deliverables, not a commercial decline. The FY2026 guidance of $18.5M reflects management estimate of when specific contract milestones will be completed and accepted by BARDA. This revenue volatility is inherent to government contract R&D businesses and should not be evaluated using the same framework applied to commercial product companies with predictable quarterly sales patterns.
The commercial hospital market is the long-term revenue opportunity that the BARDA contract is building toward. Once DeepView receives FDA clearance, the commercial market includes every emergency department, burn center, and wound care clinic in the United States. Burn injuries account for approximately 450,000 emergency department visits annually in the U.S. alone. The diagnostic gap; the 48-to-72-hour waiting period for wound depth assessment; exists in every one of those cases. A device that eliminates that waiting period and provides objective quantitative assessment has a clear clinical value proposition. The commercial pricing model, hospital procurement process, and insurance reimbursement pathway will determine how quickly the technology converts from a government-funded development program to a commercially scaled medical device business.
The cash trajectory tells a constructive story independent of the revenue fluctuations. Cash grew from $3.7M at December 31, 2024 to $10.5M at September 30, 2025 to $15.4M at December 31, 2025. The BARDA contract is not just funding the technology development; it is building the company cash position simultaneously. A medical device company that enters the commercial market with $15M-plus in cash and an established government customer relationship is in a structurally stronger position than a company that needs to raise dilutive equity to fund the commercial launch. The cash position at FDA clearance will determine how much leverage Spectral AI has in negotiating commercial distribution partnerships versus being forced to accept unfavorable terms due to capital constraints.
The military deployment pathway is the near-term commercial validation opportunity that exists independently of the FDA De Novo timeline. Military medical systems operate under different regulatory frameworks than civilian healthcare, and the BARDA contract specifically funds development of DeepView for military burn care applications. A multi-site military deployment; at combat support hospitals, military treatment facilities, or field medical stations; would simultaneously validate the technology through the most demanding institutional customer in the world and generate incremental revenue above the BARDA contract baseline. Military medical procurement decisions are made on operational need and clinical effectiveness, not on commercial sales cycles or insurance reimbursement negotiations. A successful military deployment becomes the strongest possible reference case for subsequent civilian hospital adoption because no hospital administrator can dismiss a technology that the U.S. military has evaluated and chosen to deploy.
The competitive landscape for AI-powered burn assessment is essentially nonexistent at the current stage, which is both an advantage and a risk. The advantage is that DeepView has no direct competitor with comparable clinical data, regulatory progress, or government validation. The risk is that a well-funded competitor could enter the market if the clinical and commercial opportunity becomes more visible. However, the BARDA contract itself creates a substantial barrier to entry; a competitor would need to develop their own imaging technology, generate their own clinical evidence, and win their own government contract or find alternative funding for the development program. The head start that Spectral AI has built through the BARDA relationship, the accumulated clinical data, and the submitted FDA application would take a competitor years to replicate even with unlimited funding.
The intellectual property position underpinning DeepView is built on years of clinical imaging data that cannot be replicated quickly. Machine learning models are only as good as the data they are trained on, and Spectral AI has accumulated what is likely the largest proprietary dataset of multispectral burn wound images paired with clinical outcomes in existence. This dataset was built through years of BARDA-funded clinical studies across multiple sites and patient populations. A competitor entering the market would need to generate their own training dataset from scratch; a process that requires institutional review board approvals, clinical site partnerships, patient enrollment, and years of follow-up to confirm outcome correlations. The data asset is the true moat, not the imaging hardware or the software architecture, both of which could theoretically be replicated with sufficient engineering resources. The data cannot be replicated without time, and time is the one resource that no amount of funding can compress beyond a certain point.
The addressable market sizing for AI-powered burn wound assessment extends beyond the initial military and civilian burn center applications. Chronic wound assessment; including diabetic ulcers, pressure injuries, and venous stasis wounds; represents a much larger patient population that faces similar diagnostic challenges. If the DeepView imaging platform and AI assessment capabilities can be validated for wound types beyond burns, the total addressable market expands by an order of magnitude. Chronic wounds affect approximately 6.5 million Americans annually and cost the healthcare system over $25 billion per year. The imaging and AI architecture developed for burn assessment could potentially be retrained on chronic wound datasets to address this adjacent market. This optionality is not reflected in the current valuation because it depends on future clinical validation, but the technical foundation; multispectral imaging plus machine learning classification; is applicable across wound types.
The regulatory and reimbursement pathway for DeepView in the civilian market will determine the commercial economics once FDA clearance is obtained. Medical device reimbursement in the U.S. requires either an existing CPT code that covers the procedure or the creation of a new code through the AMA CPT Editorial Panel process. Burn wound assessment using AI-powered imaging is a novel procedure category, which means Spectral AI will likely need to pursue a new CPT code; a process that takes 12 to 24 months after FDA clearance. During that period, the company can sell the device to hospitals and burn centers, but reimbursement from Medicare and private insurers may be limited to miscellaneous codes with lower payment rates. The military channel and the BARDA contract provide revenue during this reimbursement development period.
The revenue timing dynamics are critical to understanding why FY2025 revenue declined from FY2024 without any deterioration in the business. BARDA contracts pay on deliverables; specific technical milestones, documentation requirements, and demonstration benchmarks. These milestones are not evenly distributed across calendar quarters. A quarter with multiple milestone completions generates higher revenue than a quarter focused on development work between milestones. FY2024 revenue of $29.6M reflected a milestone-heavy period. FY2025 revenue of $19.7M reflected a development-intensive period. The FY2026 guidance of $18.5M reflects management expectation for the milestone cadence in the coming year. None of these fluctuations indicate a change in the total contract value or BARDA commitment.
Q4 2025 gross margin of 39.8 percent is notable for a government contract R&D company. Many BARDA-funded programs operate at cost-plus margins in the 10 to 15 percent range. A 39.8 percent gross margin suggests that Spectral AI contract structure includes meaningful margin above direct costs, which indicates BARDA confidence in the program value and Spectral AI negotiating position. The margin fluctuates by quarter based on the specific milestones completed; some milestones have higher direct costs than others. But the average margin across the contract period provides the financial foundation for ongoing operations.
The additional $31.7M BARDA award in 2025, on top of the original $150M contract. Signals continued government confidence in the program. BARDA does not increase funding for programs that are failing to meet technical milestones. The supplemental award indicates that DeepView development is proceeding satisfactorily and that BARDA sees expanded applications or accelerated timelines that justify additional investment. For investors, the supplemental award is third-party validation from the most technically demanding customer in the medical device market.
Dossiers, weekly roundup editions, and podcast notes. Free. No paywall. Unsubscribe anytime.