Fly.io has recently expanded their edge computing platform to include GPU services, positioning themselves as a solution for developers who need both edge deployment and GPU computing capabilities. The platform integrates GPU resources seamlessly with their existing infrastructure, allowing developers to deploy GPU-accelerated applications across their global network of data centers .

 

Their GPU offering is designed with a developer-first approach, featuring automatic resource scaling and per-second billing that helps optimize costs. The platform's GPU services are particularly notable for their auto-shutdown feature that terminates inactive instances after approximately 10 minutes, potentially saving significant costs for intermittent workloads. With GPU prices around $3.5/hour and flexible deployment options, Fly.io is targeting developers who need GPU resources but don't want to manage complex infrastructure or commit to long-term contracts .