Gimlet Labs

Gimlet Labs Secures $80M in Series A Funding to Accelerate AI Inference

Gimlet Labs' Series A Funding RoundGimlet Labs, an inference cloud designed to run agents, has successfully raised $80 million in a Series A funding round. This significant investment was led by Menlo Ventures and joined by Eclipse, Factory, Prosperity7, and Triatomic. The company, which launched just five months ago, has already seen its customer base triple, including a top frontier lab and a hyperscaler.Meeting the Demand for AI InferenceGimlet Labs was founded with the belief that AI inference would become a defining infrastructure challenge of the decade. This challenge has become more urgent with the rise of agentic workloads, leading to what the company refers to as the 'Inference Speed Wars.' Model providers are investing heavily in reducing agent latency, as agents are increasingly critical in workflows and produce significantly more tokens than traditional chat models.The Need for Heterogeneous InfrastructureThe AI industry is facing a wall with the current homogeneous infrastructure, which is not viable for the high-throughput, low-latency agent serving that is required. Gimlet Labs has designed its infrastructure with hardware heterogeneity at its core, building a software stack that orchestrates complex agentic workloads across heterogeneous hardware and a new kind of data center to physically connect the hardware together.This approach has resulted in 3-10X speedups on >1T parameter frontier models with large context windows, within the same power envelope. The company is also building a new type of datacenter that connects accelerators that have never coexisted before over high-speed networks, addressing challenges in connecting hardware not originally designed to communicate and managing different thermal profiles across mixed architectures.Looking AheadAs agents grow and become more complex, the need for heterogeneous infrastructure will only increase. Gimlet Labs is excited to develop the infrastructure that will make these agents faster and more useful, enabling them to do things that seemed impossible just a few years ago.