Palladyne AI Raises Low End of 2026 Revenue Forecast to $25 Million
Key Facts
On November 10, Palladyne AI raised the low end of its 2026 revenue forecast to $25.0 million, keeping the high end at $27.0 million, after reporting results for the third quarter ended September 30. The new range implies expected annual growth of between 377% and 415% compared with 2025 revenue of $5.2 million, according to the company’s statement. The previous low end of guidance was $24 million, while the high end remained $27 million. The record annual revenue adjustment puts quarterly results in the context of expectations for the rest of the year. The company reported that third-quarter revenue reached a record, alongside higher spending and losses.
Revenue was $8.6 million, up 895% year over year and 48% from the second quarter, according to the company. On a quarterly basis, component business revenue rose 26%, and management said the first delivery of next-generation BRAIN FC-1 flying computers contributed to the growth. Government contract revenue also increased 135% from the second quarter. The company said the figures reflect broader revenue growth across its component business and government contracts during the period. The company’s new annual guidance implies fourth-quarter revenue growth of between 329% and 449% compared with the same period in 2025.
Revenue growth came alongside wider losses compared with the third quarter of 2025. Operating loss was $13.4 million, versus $8.1 million in the year-earlier period, which the company attributed mainly to higher stock-based compensation, increased headcount to support growth, and higher research and development costs related to development of Gremlin-X. GAAP net loss was about $12.0 million, compared with $3.7 million a year earlier. Thus, the sharp increase in sales was accompanied by higher expenses and reported losses during the quarter. The results highlight costs alongside rapid revenue expansion.
Backlog was $18.7 million as of September 30, including $15.3 million funded, and Palladyne AI expects to recognize most of it as revenue over 12 to 18 months. The balance fell 24% during the quarter, as recognition of about $8.6 million in revenue was partly offset by approximately $2.7 million in new net contracts. The company attributed fluctuations in the balance to the timing of customers’ use of remaining funding in existing contracts before renewing them, and to continued uncertainty around defense budgets. It said, however, that the balance remained materially above its level at the start of the first and second quarters. The funded portion of backlog provides an indication of existing contracts, while the timing of renewals and revenue recognition remains tied to the factors cited by the company.
The company said cash, cash equivalents and marketable securities totaled $47.0 million at the end of September, and it expects this to cover operating, working capital and capital expenditure needs for at least 12 months from the date of the statement. At the same time, it expected full-year operating cash burn to be near or slightly above the high end of its previous range of $32.0 million to $36.0 million. Capital expenditure was about $0.9 million in the third quarter. The company added that fourth-quarter cash burn would be below the average quarterly burn since the start of the year. PDYN shares closed at $2.86 on November 18, according to EL7 data.
On commercial development, the chief executive said the U.S. Army gave the SwarmOS system the highest possible rating across the nine performance criteria it assessed under Project Convergence Capstone 6. The company included the result in its third-quarter earnings statement, released on November 10, alongside updated revenue guidance. The annual guidance move remained the announcement’s main highlight, as the company raised the low end while keeping the high end at $27.0 million. The company’s other specified figures include expected implied fourth-quarter revenue growth of between 329% and 449% year over year. It also expects to recognize most of its backlog over 12 to 18 months, according to its estimate.