Cerebras, Gimlet Labs Unveil Ultrafast AI Inference Solution

SAN FRANCISCO and SUNNYVALE, Calif., Sept. 28, 2026 /CourierPR/ -- Gimlet Labs and Cerebras Systems, a leading provider of AI infrastructure, today announced a collaboration to deliver a new class of ultrafast AI inference at massive scale. The partnership brings together Cerebras' wafer-scale compute with Gimlet Labs' Gimlet Cloud to create a purpose-built disaggregated inference cloud that spans datacenter infrastructure to developer APIs. Together, the companies plan to deliver speeds of up to 3,000 tokens per second for demanding applications, including real-time applications such as voice and video AI, agents, and assistants. The first Cerebras-powered Gimlet Cloud datacenter is expected to come online later this year.

In the world of artificial intelligence (AI), speed is a critical factor that drives user experience and engagement, and shapes what users can build. For real-time applications, such as voice and video AI, agents, and assistants, latency can be the difference between an interaction that feels seamless and one that feels slow. Real-time AI feels like an active collaborator that is immediate, fluid, and responsive. When AI responds in real time, users do more with it, stay longer, and run higher value workloads. As a result, fast tokens are more valuable tokens.

Gimlet Cloud combines Cerebras' wafer-scale compute with GPUs into an integrated inference solution. It uses advanced inference disaggregation technology to orchestrate model execution, ensuring that each phase of inference runs on the silicon best suited to it. Together, these capabilities deliver a new class of fast inference optimized for production-scale agent and real-time workloads.

The collaboration builds on joint customer engagements that have been underway since last year and an integrated solution that is already serving tokens in private deployments. Gimlet Labs will expand the collaboration to integrate software, infrastructure design, APIs, developer tooling, optimization, validation, and production operations to make ultrafast inference broadly available through Gimlet Cloud.

Gimlet Labs is a company dedicated to driving breakthrough improvements in AI performance that result in massive increases of compute available for AI workloads. Its inference cloud is derived from its foundational research across the stack, enabling the next generation of performant, scalable AI infrastructure. Its research combines theory and practice to push the boundaries of AI efficiency via techniques such as automated GPU kernel generation, workload orchestration, and heterogeneous execution across diverse hardware. Gimlet Labs is backed by Andreessen Horowitz and Menlo Ventures and is headquartered in San Francisco.

Cerebras Systems is a company that builds the world's fastest AI infrastructure. The company's team of pioneering computer architects, computer scientists, AI researchers, and engineers have come together to make AI blisteringly fast through innovation and invention. Cerebras believes that when AI is fast, it will change the world. Leading global corporations, research institutes, and governments choose Cerebras to run their AI workloads. Cerebras solutions are available on premises and in the cloud.

The collaboration builds on joint customer engagements that have been underway since last year and an integrated solution that is already serving tokens in private deployments. Gimlet Labs will expand the collaboration to integrate software, infrastructure design, APIs, developer tooling, optimization, validation, and production operations to make ultrafast inference broadly available through Gimlet Cloud.