Cerebras raises 2026 revenue forecast on AI inference demand

Cerebras Systems has raised its full-year 2026 revenue and gross margin guidance, citing strong demand for its wafer-scale AI chips as companies expand data-center infrastructure to support inference workloads.

The Sunnyvale, California-based company now expects adjusted revenue between $880 million and $890 million for 2026, up from its prior forecast of $855 million to $865 million, according to details first reported by Reuters. Cerebras also lifted its adjusted gross margin outlook to 41–43%, compared with an earlier range of 38–41%. The updated margin forecast exceeds the average analyst estimate of 35.89%.

Despite the raised targets, Cerebras shares fell more than 7% in after-hours trading following the announcement, even as the stock had gained 15.5% earlier in the week.

Second-quarter results show momentum

Cerebras reported second-quarter revenue of $180.11 million, a 74.3% increase year-over-year. The company's adjusted loss narrowed to $6.91 million from $40.5 million in the same period a year earlier.

Core cloud and services revenue—which reflects the ramp-up of its OpenAI partnership—nearly quadrupled to $127.73 million in the quarter. Finance chief Bob Komin said the company has $25.4 billion in remaining performance obligations, representing contract revenue expected to be recognized in the future, and plans to more than triple revenue in 2027.

Competing on inference with a different architecture

Cerebras is positioning itself as an alternative to Nvidia in the AI processor market by focusing on inference, the compute-intensive process that occurs when users query AI models like chatbots. The company's flagship wafer-scale engine (WSE) is a single chip the size of a dinner plate containing trillions of transistors.

This design differs fundamentally from Nvidia's approach of connecting thousands of smaller graphics processors. By placing memory directly on the chip, Cerebras says the WSE accelerates inference and reduces data-transfer delays associated with conventional GPUs that rely on separate high-bandwidth memory.

Why it matters

Cerebras is racing to scale chip production to support a $20 billion multi-year agreement to provide AI compute to OpenAI, a deal seen as critical to validating the company's market position and valuation. The raised guidance and narrowing losses suggest the company is making progress on execution, but investor caution—reflected in the after-hours stock decline—indicates the market is closely watching whether Cerebras can sustain growth and deliver on its ambitious revenue targets as it challenges an entrenched incumbent.

The financial results and updated guidance were reported by Reuters, with reporting by Anhata Rooprai in Bengaluru and Max A. Cherney in San Francisco.