AI Blocks › Ship and host your app
Cloud GPU job runner
Give it your Python code. Get it running on rented GPUs, or live as a web API, with no server to manage.
Cloud GPU job runner is a free AI skill for the job "Ship and host your app". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by Orchestra Research
Made by Orchestra Research. See the original on GitHub; we copied it as it is, with its licence.
What it does
Your Python code runs on rented GPUs, or goes live as a web endpoint, and you do not manage a server. You pay by the second, and everything shuts down to zero when idle. The setup uses Modal. You describe the container, the GPU type and the function in Python, with no config files. You can run a one-off job, a batch of many jobs in parallel, a scheduled task or a web API that scales up with traffic. The skill covers picking a GPU from small to very large, caching model files so starts are fast, storing weights and results, and keeping secrets safe. It also notes when another service suits you better, such as long-running machines.
You give it
- Your Python code or model
- The GPU size you need
- Whether it is a job, a schedule or an API
- A Modal account
You get
- A Modal app in Python
- GPU and container settings
- A deploy or run command
- Caching for model files
- Notes on cost
Use it when
- Your laptop cannot run the model you want to test.
- You want to serve a model as an API without a server.
- You have thousands of items to process in parallel.
- You need a job to run on a schedule on a GPU.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run commands and use a Modal account (paid).