Tempus AI, Inc. (Nasdaq: TEM) is a Chicago-headquartered healthcare technology company focused on applying artificial intelligence, molecular diagnostics and large-scale clinical data to precision medicine. Founded in 2015 by Eric Lefkofsky, who remains Founder and Chief Executive Officer, Tempus has developed a business that combines patient testing with a growing multimodal healthcare dataset and software intended to support clinical decisions, pharmaceutical research and drug development.
The company’s model is broader than conventional genomic testing. Tempus wants each diagnostic interaction to contribute, where permitted under applicable privacy, consent and contractual frameworks, to a larger information system linking molecular characteristics with clinical histories, pathology, treatment decisions and patient outcomes. That information can subsequently support additional diagnostics, research collaborations, artificial intelligence models and analytical tools sold to life-sciences companies.
The commercial scale is already substantial. Tempus reported second-quarter 2026 revenue of $382.5 million, up 22% year over year, with Diagnostics contributing $289.3 million and Data and Applications generating $93.2 million. Oncology testing volume increased 31%, Data Licensing and Modeling revenue grew strongly, and the company signed approximately $200 million of new Data and Applications licences during the quarter.
Growth has increasingly been accompanied by acquisitions. Ambry Genetics expanded Tempus into hereditary testing, Deep 6 AI strengthened clinical-trial matching, Paige added artificial intelligence-powered digital pathology and OneOme broadened the company into pharmacogenomics. The proposed acquisition of Personalis would go further by bringing the technology and laboratory operations underlying Tempus’ existing molecular residual disease offering under common ownership.
That expansion creates a distinctive company profile but also raises the execution bar. Tempus now has to prove that diagnostics, healthcare data and artificial intelligence can reinforce one another economically rather than simply coexist inside a growing collection of businesses. Sustainable profitability, reimbursement, acquisition integration and recurring pharmaceutical demand will determine whether the platform becomes more valuable as it becomes larger.
What does Tempus AI actually do across diagnostics, data and artificial intelligence?
Tempus increasingly reports its business through two broad categories, Diagnostics and Data and Applications. The categories are financially separate but strategically interconnected because diagnostic testing generates clinical and molecular information that can increase the value of Tempus’ datasets and analytical capabilities.
Diagnostics includes molecular testing used primarily in oncology and hereditary disease, with additional expansion into areas such as pharmacogenomics and cancer monitoring. The company’s oncology products are intended to help physicians understand the molecular characteristics of a patient’s tumour, identify potentially relevant therapies and, increasingly, monitor disease over time.
Ambry Genetics substantially widened this business after becoming a wholly owned Tempus subsidiary in February 2025. Ambry brought a large hereditary-testing operation into a company whose diagnostics identity had previously been associated much more closely with oncology, giving Tempus broader access to inherited-disease data and clinical relationships.
Data and Applications monetises the information infrastructure surrounding those diagnostic activities. Tempus provides pharmaceutical and biotechnology companies with access to de-identified multimodal datasets, analytical services, data licences, artificial intelligence models and technology supporting clinical research and drug development.
The interaction between the two businesses is central to the strategy. A diagnostic test creates revenue when performed, but the resulting information can potentially create further economic value when incorporated into appropriately governed datasets used for research, modelling and future product development. Tempus is therefore attempting to earn at several points in the healthcare information chain rather than relying exclusively on reimbursement for individual laboratory tests.
Why is Tempus’ multimodal healthcare data library central to the business model?
Artificial intelligence in healthcare depends heavily on access to high-quality data that are both sufficiently large and sufficiently contextualised. Hospitals, laboratories, imaging systems, electronic medical records and pharmaceutical companies often hold different fragments of the same patient journey, making it difficult to analyse how molecular characteristics relate to treatment decisions and eventual outcomes.
Tempus has spent much of its history assembling and organising these different information types. Its datasets can include molecular data, clinical histories, treatment information, pathology and other medical variables, allowing researchers to analyse relationships that may be difficult to identify when each data source is considered separately.
The commercial attraction is straightforward. A conventional diagnostics laboratory may earn revenue each time it performs a test, whereas Tempus can potentially generate additional revenue by licensing de-identified datasets, providing modelling services or helping pharmaceutical companies identify biomarkers and patient populations for drug development.
This creates the possibility of a reinforcing loop in which more diagnostics produce more useful data, richer data improve analytical products, and stronger analytical products attract more pharmaceutical and clinical customers. The model becomes economically powerful only if Tempus can continue adding high-quality information while preserving privacy, consent, data rights and customer trust.
The size of the dataset alone therefore does not create a durable competitive advantage. Tempus must demonstrate that combining different healthcare modalities produces insights or workflows valuable enough for physicians and life-sciences companies to keep paying for them.
How did Ambry Genetics change Tempus AI’s diagnostics business?
Tempus completed its acquisition of Ambry Genetics on February 3, 2025, paying $375 million in cash and $225 million in Tempus shares at closing. Ambry continued operating as a wholly owned subsidiary and brought an established hereditary-testing franchise into the broader Tempus platform.
The acquisition materially increased the scale of Diagnostics. Full-year 2025 Diagnostics revenue reached approximately $955 million, with the year-on-year increase reflecting both continued oncology growth and the addition of Ambry’s hereditary-testing business.
By the second quarter of 2026, Ambry was fully reflected in both comparative periods, making underlying hereditary growth easier to evaluate. Hereditary revenue increased approximately 5% to $107.4 million, while hereditary testing volume grew around 2% to approximately 141,500 tests.
Those figures were considerably slower than oncology, where testing volume increased 31%. The contrast matters because it shows that Ambry should not automatically be treated as another high-growth engine simply because it enlarged Tempus dramatically at acquisition.
The strategic value is nevertheless broader than one quarter’s growth rate. Hereditary testing can identify inherited disease risk before or independently of an active cancer diagnosis, potentially extending Tempus’ relationship with healthcare providers and patients into different stages of care.
The longer-term question is whether Tempus can use Ambry’s clinical relationships and genomic information to strengthen the wider data platform while restoring stronger organic growth in hereditary testing. If Ambry becomes primarily a source of acquired scale without meaningful cross-platform benefits, the economics will be less attractive than the strategic narrative suggests.
What does OneOme add to Tempus through pharmacogenomics?
Tempus expanded further into personalised medicine through its acquisition of OneOme in November 2025 and the subsequent national launch of its pharmacogenomics offering in July 2026.
OneOme’s RightMed Comprehensive test analyses 27 genes associated with medication response, while targeted testing includes genes such as DPYD and UGT1A1 that can be relevant when assessing how particular patients may metabolise or tolerate specific drugs.
The commercial logic differs from tumour sequencing but fits the same underlying precision-medicine thesis. Genomic information can help clinicians understand not only characteristics of a disease but also how an individual patient may respond to treatment, potentially influencing medication selection, dose or toxicity management.
Pharmacogenomics therefore extends Tempus beyond identifying what disease a patient has and which therapies may target it. The company is increasingly attempting to provide information about how a particular patient may respond once therapy begins.
OneOme also adds another genomic data type to the broader Tempus ecosystem. Whether that becomes strategically meaningful will depend on physician adoption, reimbursement and evidence that pharmacogenomic testing changes treatment decisions often enough to justify wider use.
Why does the proposed Personalis acquisition matter for molecular residual disease?
Tempus announced a definitive agreement in July 2026 to acquire Personalis, Inc., a precision oncology company specialising in tumour-informed molecular residual disease technology. The transaction values the remaining Personalis business at approximately $1.5 billion on an enterprise-value basis net of Tempus’ existing ownership interest, with Personalis shareholders entitled to receive consideration valued at $16.25 per share.
The agreement is structured as a 100% stock transaction, although Tempus has discretion to substitute cash for as much as 50% of the consideration. Completion is expected in late 2026 or early 2027 and remains subject to Personalis shareholder approval, regulatory clearances and customary closing conditions.
The transaction builds on an existing relationship rather than taking Tempus into molecular residual disease for the first time. Tempus invested in Personalis and became the exclusive commercial distributor of NeXT Personal in 2023, while Personalis retained responsibility for the underlying assay and laboratory testing.
NeXT Personal is designed to detect very small amounts of circulating tumour DNA following treatment, potentially allowing clinicians to identify residual disease, monitor treatment response or detect recurrence earlier than conventional methods.
Tempus reported approximately 9,000 molecular residual disease tests during Q2 2026, up from around 6,500 in the first quarter, as it continued commercialising NeXT Personal. Personalis separately reported 10,384 total clinical tests during the quarter, but those figures should not be added together because Personalis performs NeXT Personal tests generated through the existing Tempus commercial relationship.
The transaction is therefore less about combining two independent MRD testing businesses than about bringing the technology, laboratory operations, reimbursement economics and commercial distribution of an existing partnership under common ownership.
If completed successfully, the acquisition could give Tempus greater control over the economics and data generated across a patient’s cancer journey. Initial tumour profiling may help determine treatment, while repeated MRD testing can create a longitudinal relationship extending through therapy and post-treatment monitoring.
That recurring element is particularly attractive strategically, but the opportunity remains dependent on clinical adoption and reimbursement. Tempus will need to demonstrate that greater ownership of the MRD platform translates into economically attractive testing growth rather than simply adding another large acquisition to an already complex organisation.
How are Deep 6 AI and Paige expanding Tempus beyond molecular testing?
Tempus has also used acquisitions to extend its artificial intelligence capabilities into clinical research and digital pathology.
Deep 6 AI, acquired in March 2025, provides software that analyses structured and unstructured information from electronic medical records to identify patients who may be eligible for clinical trials. At the time of the acquisition, the platform was integrated with more than 750 provider locations covering over 30 million patients.
Clinical-trial recruitment fits naturally with Tempus’ broader data strategy because finding appropriate patients remains one of the most persistent bottlenecks in drug development. Combining molecular characteristics with current clinical information could potentially help pharmaceutical companies identify eligible patients more efficiently while giving health systems additional research opportunities.
Paige added another data modality when Tempus acquired the digital-pathology artificial intelligence company in 2025. Pathology slides contain enormous quantities of visual biological information, and digitising those images creates the possibility of analysing them alongside genomic results, treatment histories and clinical outcomes.
The potential strategic benefit is not simply selling another pathology product. Connecting digital pathology with molecular and clinical data could allow Tempus to build models using a richer representation of disease than any single modality provides.
This is increasingly what distinguishes the Tempus acquisition strategy. Ambry adds hereditary genetics, OneOme adds pharmacogenomics, Deep 6 contributes clinical-trial matching, Paige brings pathology and Personalis could deepen cancer monitoring. The challenge is making those capabilities operate as parts of one information platform rather than allowing the company to become a collection of loosely integrated healthcare assets.
Can Tempus turn foundation models into a real pharmaceutical research business?
Tempus’ foundation-model strategy is being developed partly through its collaboration with AstraZeneca and Pathos AI. Under the contractual structure, Pathos is responsible for development activities for the oncology foundation model, while Tempus contributes de-identified multimodal healthcare data and supporting infrastructure and is responsible for providing the resulting model to AstraZeneca.
Tempus reported during Q2 2026 that the first version of the oncology foundation model had been successfully delivered to AstraZeneca.
The collaboration illustrates why Tempus’ data library could have value beyond conventional diagnostics. Pharmaceutical companies spend enormous amounts on drug development while many programmes fail because suitable biological targets or responsive patient populations are difficult to identify.
Models trained across molecular and clinical data could potentially help researchers identify biomarkers, understand disease subtypes, select trial populations or generate hypotheses that would be difficult to derive from smaller or isolated datasets.
Evidence of commercial demand can already be seen in Data and Applications. Tempus signed approximately $200 million of new licences during Q2, including agreements involving companies such as BioNTech, Daiichi Sankyo, Level Set Bio and Incyte Pharmaceuticals.
Data and Applications revenue reached $93.2 million for the quarter, representing 28% year-over-year growth, while Insights revenue grew 36%.
The commercial significance of foundation models should nevertheless be judged by repeat usage and durable revenue rather than by partnership announcements. Pathos’ role also makes it important not to portray Tempus as solely responsible for model development. Tempus’ distinctive contribution lies primarily in its data, infrastructure and ability to connect model development with a large precision-medicine ecosystem.
What does FDA approval of tumor-only xT CDx change for Tempus?
Regulatory progress is becoming increasingly important as Tempus moves parts of its diagnostics portfolio from laboratory-developed testing toward products carrying formal United States Food and Drug Administration approval.
During 2026, the Food and Drug Administration approved an expanded tumor-only configuration of Tempus’ xT CDx assay. The 648-gene test can now be used without a matched blood or saliva specimen when that matched normal sample is unavailable.
Tempus says the approval makes it the first laboratory with companion-diagnostic approval for both tumor-only and tumor-normal comprehensive genomic profiling.
The distinction has practical implications because obtaining a matched normal specimen is not always convenient or possible. Tumor-only testing can reduce that barrier while preserving access to a broad molecular profile that may help identify clinically relevant biomarkers and determine eligibility for particular targeted therapies.
The approval also carries economic implications. Tempus has indicated that migration toward unified Advanced Diagnostic Laboratory Test pricing could increase average selling price by approximately $200 and potentially contribute around $85 million of annualised revenue beginning in 2027.
That estimate is management’s expectation rather than guaranteed future revenue, but it demonstrates how regulatory progress can affect both clinical adoption and reimbursement economics.
Formal approval also increases compliance obligations. Tempus must continue satisfying applicable quality, validation, manufacturing and post-market requirements, meaning regulatory advancement strengthens commercial credibility while making the operating environment more demanding.
What do Tempus AI’s Q2 2026 results reveal about underlying profitability?
Tempus reported GAAP net income of $5.6 million during the second quarter of 2026 compared with a $42.8 million net loss a year earlier, but the headline result materially overstates the profitability of the underlying operations.
The quarter included approximately $98.5 million of unrealised gains on marketable securities. Tempus simultaneously recorded $55.6 million of stock-based compensation expense and related employer payroll taxes, while the actual business generated a GAAP operating loss of approximately $75.9 million compared with an operating loss of around $61.8 million a year earlier.
Q2 should therefore not be described simply as Tempus becoming profitable. The company achieved positive net income because investment gains more than offset continuing operating losses.
Adjusted EBITDA provides a more encouraging measure of underlying progress. Tempus generated positive adjusted EBITDA of $8 million compared with a $5.6 million adjusted EBITDA loss in the corresponding period, marking an important improvement in operating economics even though sustainable GAAP operating profitability has not yet been established.
Gross profit increased 26% to $246.5 million, while consolidated gross margin improved from approximately 62.0% to 64.4%. Diagnostics gross margin strengthened from 58.8% to 62.6%, reflecting improving testing economics.
Data and Applications remained the structurally higher-margin business with a 70.2% gross margin, but that figure actually declined from 72.7% a year earlier as cloud and foundation-model development costs increased. The distinction matters because faster Data and Applications growth could improve Tempus’ overall economics, but only if the costs associated with delivering those products remain controlled.
The profitability story is therefore moving in the right direction without being complete. Tempus has demonstrated positive adjusted EBITDA at meaningful scale, but the company still needs operating losses to narrow before sustainable accounting profitability can be considered established.
What does Tempus’ 2026 financial outlook say about its growth trajectory?
Following its second-quarter performance, Tempus increased full-year 2026 revenue guidance to between $1.595 billion and $1.605 billion, implying approximately 25% annual growth. Management maintained its expectation for around $65 million of adjusted EBITDA.
The guidance excludes contribution from the proposed Personalis acquisition because the transaction is expected to close only in late 2026 or early 2027. That makes the current outlook a clearer measure of the growth expected from existing Tempus operations.
The company has also presented an illustrative long-term framework under which revenue could grow approximately 25% annually over the next three years, supported by continued Diagnostics expansion and increasing scale in Data and Applications. Tempus explicitly states that this framework does not constitute formal financial guidance.
The distinction is important because multi-year healthcare growth depends on variables including reimbursement, clinical adoption, pharmaceutical bookings and regulatory progress that can change materially over time.
The composition of revenue growth will matter as much as the headline percentage. Diagnostics supplies scale and data, but faster growth in Data and Applications could improve margins and demonstrate that Tempus is monetising its information platform after the initial clinical interaction.
Oncology testing remains one of the strongest current engines, while hereditary testing needs to demonstrate stronger organic momentum. Pharmaceutical licensing and foundation-model relationships will also need to develop into recurring commercial activity rather than depending excessively on periodic large contracts.
How strong is Tempus’ balance sheet as acquisitions expand?
Tempus ended June 2026 with approximately $820.7 million of cash and marketable securities, giving the company substantial liquidity for operations and strategic investment.
That liquidity should be viewed alongside a broader debt structure rather than only the most recent financing. Tempus completed a $460 million offering of 0.00% convertible senior notes due in 2032 during the second quarter, while it also had $750 million principal amount of 0.75% convertible senior notes due in 2030 and approximately $187.9 million of convertible promissory-note liabilities at June 30.
The company had repaid borrowings under its revolving credit facility and remaining term-loan principal by May 2026, simplifying part of the capital structure even as convertible obligations increased.
Zero-coupon and low-coupon convertible debt reduce near-term cash-interest requirements compared with conventional borrowing, but they still create future repayment obligations and potential shareholder dilution depending on conversion conditions and future equity prices.
These financing considerations become particularly relevant because Tempus has completed several acquisitions within a short period and could use cash for up to half of the Personalis merger consideration.
Tempus has stated that existing liquidity should be sufficient to finance its current operating plan for at least one year from issuance of its Q2 report. The more important long-term question is whether acquisitions increasingly fund themselves through stronger cash generation rather than requiring repeated external financing.
Capital allocation therefore deserves close attention. Ambry, Deep 6 AI, Paige and OneOme have widened the platform, while Personalis would be a substantially larger transaction. Each deal can strengthen Tempus strategically, but every acquisition also brings integration costs, goodwill and the possibility that anticipated synergies do not materialise.
What are the biggest risks facing the Tempus AI business model?
Reimbursement is one of the most fundamental risks because much of Diagnostics ultimately depends on payment from Medicare, commercial insurers and other healthcare payers. Clinical utility does not automatically guarantee favourable coverage, and changes in reimbursement policies can materially affect revenue per test.
Regulatory exposure is similarly broad. Tempus operates clinical laboratories, develops diagnostic products and increasingly pursues formal Food and Drug Administration approvals, creating obligations spanning laboratory regulation, medical devices, clinical validation, privacy and data security.
Personalis would increase that complexity by bringing additional cancer-monitoring technology and laboratory operations inside Tempus.
Healthcare data creates another particularly sensitive risk. Much of Tempus’ strategic value depends on assembling and analysing clinical and molecular information, meaning privacy, consent, cybersecurity and contractual rights are essential to the business model. A serious security breach or misuse of data could damage both customer trust and regulatory standing.
Acquisition integration may become one of the largest strategic risks as the company expands. Ambry Genetics, OneOme, Deep 6 AI, Paige and the proposed Personalis transaction give Tempus access to a wide range of valuable technologies, but assembling assets is substantially easier than creating one coherent commercial platform.
Competition is also fragmented across several categories. Tempus competes with genomic-testing laboratories, precision-oncology businesses, clinical research technology companies, healthcare data providers and artificial intelligence platforms. Its strategic defence is the combination of these capabilities, but that advantage exists only if integration produces better products or economics than specialist competitors can achieve independently.
Finally, sustainable profitability remains unfinished. Positive adjusted EBITDA demonstrates progress, but continuing operating losses mean Tempus must keep improving productivity and margins while simultaneously funding product development and acquisition integration.
What will determine whether Tempus becomes a durable precision medicine platform?
Tempus has already demonstrated that diagnostic testing and healthcare data can coexist inside the same commercial model. The more difficult phase now begins: proving that greater diagnostic scale continues enriching the data business, that pharmaceutical customers convert large licensing agreements into recurring relationships and that artificial intelligence products deliver measurable research or clinical value rather than functioning primarily as an attractive technology narrative.
The Personalis transaction will be one of the most important tests because Tempus already commercialises NeXT Personal. The acquisition is therefore less about entering molecular residual disease for the first time than about internalising technology, laboratory economics and longitudinal cancer-monitoring data that currently sit outside Tempus. Successful integration could broaden the company’s role from therapy selection into repeated monitoring across a patient’s cancer journey, while reimbursement will determine how effectively increasing MRD volumes translate into attractive revenue.
Acquisition integration across the rest of the platform will matter for similar reasons. Ambry gives Tempus hereditary genetics, OneOme adds pharmacogenomics, Paige contributes digital pathology and Deep 6 AI connects the company more closely with clinical-trial recruitment. If those businesses remain largely independent, Tempus will resemble a diversified diagnostics and healthcare technology group. If their data and workflows become genuinely interconnected, the company could build something closer to a common precision-medicine infrastructure layer.
Profitability provides another measurable threshold. Tempus has reached positive adjusted EBITDA, but Q2 still produced a $75.9 million operating loss and GAAP net income benefited substantially from unrealised investment gains. Sustainable progress therefore requires operating losses to narrow as Diagnostics scales, while Data and Applications must expand without allowing cloud and model-development costs to erode the margin advantage of that business.
The pharmaceutical side of the platform also needs to prove repeatability. The AstraZeneca and Pathos foundation-model collaboration, approximately $200 million of new Q2 Data and Applications licences and relationships with companies including BioNTech and Daiichi Sankyo demonstrate commercial interest, but long-term value will depend on customers renewing, expanding and repeatedly using Tempus data and models across drug-development programmes.
The pieces of a much broader precision medicine platform are increasingly visible across oncology sequencing, hereditary testing, pharmacogenomics, digital pathology, clinical-trial matching, molecular residual disease and pharmaceutical data services. Whether those pieces ultimately create a durable competitive system will be determined by integration, reimbursement, recurring data revenue and evidence that Tempus can convert the increasing scale of its multimodal healthcare dataset into better products without allowing acquisition complexity and capital requirements to outpace the economic benefits.
Discover more from Business-News-Today.com
Subscribe to get the latest posts sent to your email.
