Cerebras Systems raised its full-year revenue forecast on November 12, 2026, after third-quarter GAAP revenue rose 55% year over year to a record $210.8 million. Cloud and other services revenue jumped 265% to $141.9 million, growth the company attributed to demand for high-speed AI inference, but its GAAP net loss widened to $161.2 million.
Cerebras now targets $940 million to $960 million in full-year core revenue, up from the $880 million to $890 million range in its August 12, 2026, second-quarter release. Its fourth-quarter outlook calls for a core operating margin between negative 1% and positive 1%, putting the adjusted measure close to breakeven. Core results exclude stock-based compensation and data-center pass-through items and add back noncash customer-warrant amortization recorded against revenue, among other adjustments.
Cloud Services Drive the Revenue Mix
Cloud and other services supplied most of the quarter’s GAAP revenue, while hardware revenue declined from a year earlier. That shifts more of Cerebras’s growth toward delivering inference capacity as a service, complementing direct system sales and access through third-party clouds.
Cerebras reported $26.1 billion in remaining performance obligations as of September 30, 2026, up from $25.4 billion at June 30, 2026. The scale of those commitments makes deployment schedules and available capacity central to turning customer demand into reported revenue.
Cerebras also announced that Oracle is integrating its fast inference into Oracle Cloud Infrastructure and that it had signed a cloud agreement with an unnamed customer it described as a top-three hyperscaler. Oracle availability is planned for 2027, and more information about the hyperscaler agreement is expected that year.
GAAP Losses and Infrastructure Spending
The GAAP net loss compares with $22.2 million a year earlier. Stock-based compensation represented $155.1 million of the adjustments in the third-quarter operating reconciliation, helping explain why Cerebras’s core operating result is much closer to breakeven than its GAAP result.
Expansion is also cash-intensive. During the first nine months of 2026, Cerebras used $388.9 million in operating activities and purchased $1.046 billion of property and equipment. It ended September with $7.8 billion in cash, cash equivalents, restricted cash and short-term investments, providing substantial resources to finance the buildout.
Capacity Delivery Becomes the Next Test
Cerebras has secured 700 gross megawatts of data-center capacity for delivery by the end of 2027, including new contracts in Norway and Finland. Its August release separately described more than 600 megawatts live or under contract for delivery by the same deadline, with that earlier figure also encompassing capacity already operating.
Manufacturing is scaling in parallel with Flex and Sanmina. Cerebras expects to exit 2026 with more than 10 times the manufacturing capacity it had at the beginning of the year, with plans for additional factories and capacity in 2027.
The company says its disaggregated inference architecture can increase the number of tokens a data center delivers by up to five times, improving the potential revenue produced per megawatt. That capability is central to its effort to supply high-speed inference without requiring a proportional increase in powered infrastructure.
The raised outlook ties Cerebras’s next phase to execution as much as demand. Converting customer commitments into deployed inference capacity while managing cash use and GAAP losses will determine whether rapid cloud growth produces sustainable compute economics.
Credit: Source link
