After a Near 50% Drop, Tempus AI Could Be Ripe for a Rebound

Key Points

• Tempus AI’s post-IPO volatility has reset expectations, but Q4 results showed strong organic growth alongside acquisition-fueled gains.
• Management’s 2026 outlook points to a potential full-year adjusted EBITDA inflection, a key milestone for the story.
• The company’s large multimodal dataset and entrenched testing channels may make disruption harder, even as analyst targets trend lower.

Since going public in mid-2024, shares of healthcare and life science services company Tempus AI NASDAQ: TEM have gone on a wild rollercoaster ride. The firm’s initial public offering price was $37. The stock went on to eclipse $85 by February of 2025, only to fall to around $50 two months later. By October, shares rebounded above $100. Notably, around this time, MarketBeat cautioned investors, as Tempus flashed multiple overvaluation signals. The stock now trades near $54 per share, dropping almost 50% from its all-time high.

Tempus just reported its latest financial results, and unfortunately for the firm, it didn’t receive much love from the market, dropping 7% afterward. Still, with shares down so drastically, this is a company worth revisiting.

TEM Tops Estimates, Projects Full-Year EBITDA Inflection in 2026

In Q4 2025, Tempus grew its revenue by 83% to $367 million, surpassing estimates. However, it is important to note that much of this growth came from acquisitions, inflating the topline figure. Still, the company’s organic growth came in at an impressive 33.5%. The company also managed to post an adjusted loss per share of 4 cents, slightly better than the 5-cent loss analysts anticipated.

Testing volume growth in the company’s oncology and heredity business lines came in at 29% and 23%, respectively. Maintaining solid growth here is key to the company's success. More tests not only generate diagnostic revenue, but also eventually feed into the company’s data and applications revenue.

Tempus sells the data generated from its tests to pharmaceutical companies that use it to boost their chances of clinical trial success. Data and applications revenue rose 25%, a slight deceleration from 26% last quarter. The company’s net retention rate was 126% for the full year, meaning that existing customers from 2024 increased their spending by 26% in 2025.

Along with strong growth, Tempus AI’s profitability is improving. Adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) totaled $12.9 million, up from -$7.8 million in Q4 2024.

Looking ahead, Tempus expects to grow revenues by a solid 25% in 2026 and forecasts adjusted EBITDA of $65 million. This would put the figure in positive territory for the first time in Tempus’s history over a full year.

TEM’s Huge 450 Petabyte Dataset: An AI-Disruption Shield

When it comes to the potential for disruption from artificial intelligence, the threat to Tempus is questionable. Notably, Tempus has a massive 450 petabytes of multimodal data that it can use to train its own models.

Conveniently, data storage company Pure Storage NYSE: PSTG, which will begin trading under the name Everpure on March 5, helps put this number into context. They estimate that with 3.6 billion exams conducted annually, the healthcare sector generates around 450 petabytes of imaging data per year. Thus, Tempus’s dataset is similar in size to all the imaging the healthcare system produces annually, a staggering statistic.

Not only is Tempus’s head start in accumulating this data huge, but it is also growing and would be difficult for emergent competitors to replicate. Tempus’s dataset has more than tripled since 2022, when it stood at less than 150 petabytes. Tempus’s data also flows from its diagnostics segment, where it performs oncological and hereditary testing. The company states that over 8,500 oncologists and thousands of other physicians are regularly ordering tests, giving it a strong foothold in this group.

Generating this training data would require AI companies to also perform this testing. This is something that none of the big names do, and likely would not have an interest in. For AI labs developing agents, working to disrupt software-as-a-service business models is a much more logical direction. Their AIs have already been trained on coding data; they don’t necessarily need to accumulate massive amounts of other data to build competing products.

Tempus is currently developing its own foundational model, which it has submitted to pharma giant AstraZeneca NASDAQ: AZN for review. This foundational model is a potential revenue source that Tempus has not significantly tapped into at this point.

Even After Lowering Targets, Analysts Eye Solid Gains Ahead

The MarketBeat consensus price target on Tempus sits near $79, implying over 40% upside in shares.

However, it is important to recognize that many analysts have not updated their targets for quite some time.

Among targets updated or issued after Tempus’s earnings release, the average is moderately lower at approximately $71. Still, this figure implies substantial upside potential of over 30%.

Overall, Tempus is clearly seeing strong demand for its offerings. Existing customers also keep spending more, a strong indication that Tempus is providing real value and may be generating upselling opportunities.

Considering all this, Tempus shares look like they have a significant opportunity to rebound.