1. Home
  2. Companies
  3. Gimlet Labs

About

Gimlet Labs is an applied research lab focused on building next-generation AI infrastructure for the agentic era. The company's work centres on making AI workloads ten times more efficient by expanding usable compute and improving orchestration. Its technical domains span heterogeneous compute, workload orchestration, serverless inference, autonomous kernel generation, and multi-agent architectures.

The company has developed two products. Gimlet Cloud is a serverless inference platform designed for AI agents. kforge is an autonomous kernel generation system that employs multi-agent architectures to explore different designs and automatically identify the fastest kernels. The infrastructure is built to support deployment to gigawatt-class AI datacenters.

Gimlet Labs operates across the AI infrastructure, cloud computing, and datacentre sectors. As an applied research lab, its stated mission is to make AI workloads 10X more efficient, a goal that shapes its product and engineering priorities.

Similar companies

Together AI logoTA

Together AI

Together AI is a research-driven AI cloud infrastructure provider enabling developers and enterprises to train, fine-tune, and deploy open-source generative AI models at scale.

WEKA logoWE

WEKA

WEKA develops AI-native data infrastructure software that accelerates machine learning and high-performance computing workloads across cloud and hardware environments.

Moonlite logoMO

Moonlite

Moonlite provides high-performance, co-located AI infrastructure and cloud-native platforms for large-scale model training and compute-intensive workloads for enterprises and research institutions.

MemVerge logoME

MemVerge

MemVerge develops memory-converged infrastructure software to optimize big memory, GPU-centric computing for AI and multi-cloud environments.

Tensorlake logoTE

Tensorlake

Tensorlake is a sandbox-native cloud platform providing isolated MicroVM environments and scheduling infrastructure for AI agents, agentic applications, and data workflows.

FriendliAI logoFR

FriendliAI

FriendliAI builds an optimized AI inference platform that accelerates and reduces the cost of running large language models through custom GPU kernels and advanced batching techniques.