crunr provides a command-line interface for running code on cloud-based GPU compute resources without requiring users to manage infrastructure or possess cloud expertise. Designed to eliminate the overhead of manual cloud setup, DevOps, and billing surprises, crunr enables users to execute local scripts in the cloud with a single command. The platform is intended for AI engineers, indie hackers, startups, data scientists, researchers, students, and developers who need access to scalable compute power for tasks such as machine learning training, batch jobs, and experiments.
The tool operates by allowing users to write scripts in any language they would typically run locally—such as Python, Node, bash, R, or Go—and then execute them in the cloud using the crunr CLI. Upon command execution, crunr automatically selects the most cost-effective matching spot instance, synchronizes code and dependencies, runs the job, and ensures that outputs are returned directly to the user's laptop. The compute instance is terminated immediately after the job completes or if it crashes, ensuring that users are only billed for active compute time and not for idle resources. This approach addresses common pain points such as idle billing, complex setup, failed deployments, and the need for ongoing monitoring.
crunr offers two modes of operation: crunr cloud and bring-your-own-AWS (BYOC). With crunr cloud, users do not need an AWS account, IAM user, or cloud policy setup; instead, they obtain an API key and can begin running jobs in under a minute. For those who prefer direct control, the BYOC mode allows users to connect their laptop directly to their own AWS account, with credentials and data remaining private and no relay through crunr's backend. Both modes are fully compatible with the crunr CLI, and the platform supports integration into shell, CI, and scripting workflows.
xlarge GPU instance and actual job costs calculated by the minute. There are no additional charges for idle time, setup, or DevOps overhead, as these aspects are automated by the tool. crunr is positioned as a solution for anyone needing to run scripts on cloud GPUs efficiently, without the traditional complexity and hidden costs of cloud infrastructure management.
crunr is a Hosting, deployment & PaaS project. It enables users to easily run and manage GPU compute jobs on cloud infrastructure or their own AWS EC2 instances. It is built as an open-source project for machine learning engineers and researchers. The project is open source (MIT). It runs on the web, the command line, and API.
sandy088 builds and maintains crunr, and it first shipped in 2026. Key capabilities include GPU job scheduling, cloud compute, and AWS EC2 integration.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match