Spectral AI enters its third-quarter report with a split personality. Strategically, the company has never looked better: the May de novo clearance of its DeepView burn system finally turned a years-long regulatory wait into a commercial launch. Financially, the last quarter was its weakest in some time, with revenue slipping and gross margin collapsing as government funding shifted into a cost-share phase. The report due after the close on November 10 is the first real test of whether the second half can close that gap.
The Street expects a loss of $0.14 per share, slightly wider than the $0.13 loss posted both last quarter and a year ago. Revenue is pegged at $5.3 million, up roughly 41% from $3.8 million in the year-ago quarter and a big jump from the $3.5 million reported in the second quarter. That sequential leap is the crux of the setup. Management has reaffirmed full-year revenue of about $18.5 million for three straight quarters, and consensus of $18.56 million sits essentially on top of it. With roughly $7.5 million booked in the first half, Spectral needs about $11 million in the back half, which implies two quarters averaging well above anything it has delivered recently. A third-quarter print near $5.3 million would make the target look credible; a figure closer to $4 million would put heavy pressure on the fourth quarter and revive memories of the cut from $21.5 million last year.
What makes this especially interesting is that the guidance explicitly excludes meaningful DeepView sales. The ramp, in other words, has to come largely from contract and development revenue, chiefly BARDA work, rather than from the commercial launch. Investors should want clarity on what is driving the step-up and whether it comes with margin repair. Gross margin fell to 31.6% last quarter from 50.8% in the first quarter, and management attributed the drop to the cost-share structure. If margin stays near that level while revenue rises, the improvement in the loss will be limited, especially with operating expenses having climbed to $5.4 million on launch preparation and stock compensation.
On the commercial side, the evidence that matters is early and qualitative but still important. The first 30 U.S. installations are subsidized under BARDA through mid-2027, so the count of signed or completed placements is a cleaner gauge of demand than revenue. Management previously described a wait list of interested sites while also flagging procurement, cybersecurity and AI-committee reviews as friction at hospitals without existing relationships. Progress on converting that list, along with hiring beyond a two-person sales team under the new chief commercial officer, would validate the launch narrative. Updates on UKCA expanded clearance, still expected this quarter, the 12-center U.S. outcome study, and any word on the undecided EmTech handheld follow-on would round out the picture. Cash of $14.0 million at midyear, plus access to more than $60 million in BARDA funding and no debt principal due until March 2027, should keep liquidity questions in the background.
The market has grown more skeptical. Bearish sentiment has deepened to about 48% from roughly 30% heading into the last report. The stock is unchanged since then, technically beating a 2.9% decline in the S&P 500, but at $1.52 it trades below its 200-day average of $1.70 and near the bottom of its post-earnings range of $1.43 to $1.81, a far narrower and lower band than last quarter's $1.50 to $2.97 swing. That suggests expectations have been reset lower, though not abandoned.
Ultimately, this report turns on whether the revenue rebound shows up on schedule. A quarter near consensus with stabilizing margins and concrete placement progress would let the clearance story carry the stock. Another soft quarter would leave Spectral with a promising product and a full-year target that looks increasingly hard to reach.