Cerebras will supply 100 megawatts of AI chip capacity to cloud startup Gimlet Labs, marking a major infrastructure deployment for an emerging neocloud competitor.
Cerebras Systems and startup Gimlet Labs announced a partnership Monday to deploy 100 megawatts of Cerebras-powered AI inference capacity through Gimlet's cloud platform. Cerebras will supply its CS-4 systems to Gimlet over one to two years, with Gimlet planning to make them available through its cloud in 2027. The first Cerebras-powered Gimlet Cloud datacenter is expected to come online later this year. The companies declined to disclose financial terms.
Gimlet's platform employs what the company calls a disaggregated inference architecture, combining Cerebras wafer-scale processors with graphics processing units. The approach routes different phases of AI inference—prefill, decode, and attention operations—to whichever type of chip is best suited for each task. Together, the companies plan to deliver speeds of up to 3,000 tokens per second for agentic and real-time applications.
"By combining Gimlet's multi-silicon software with the Cerebras Wafer Scale Engine, we can run each phase of inference on the hardware best suited to it and plan to deliver up to 3,000 tokens per second at production scale," said Gimlet Labs Co-founder and CEO Zain Asgar.
Cerebras Co-founder and CTO Sean Lie said that combining the company's chips with high-throughput GPUs improves datacenter economics. "By making Cerebras a native part of its inference cloud, Gimlet will bring our industry-leading speed and intelligent AI to more developers at production scale," he said. "We're excited to build with Gimlet as a launch partner for CS-4, giving customers a direct path to our latest technology."
According to Asgar, fast inference supports use cases including cybersecurity, voice applications, and financial analysis. Gimlet plans to target companies and startups building products around AI, and intends to run frontier models—the most capable AI systems available—on the Cerebras hardware. After delivery, Gimlet assumes responsibility for day-to-day upkeep and operation of the systems.
The partnership builds on joint customer work conducted since last year, with an integrated solution already handling inference traffic in private deployments. Gimlet is backed by Andreessen Horowitz and Menlo Ventures.
Cerebras reported second-quarter core revenue of $210 million in August and raised its full-year outlook. The company has signed similar supply agreements with other major AI customers, including a deal earlier this year with OpenAI.