Skip to content

Gridweave

Scale workloads from your notebook 🚀

Run your code on remote machines directly from your notebook. Launch parallel jobs, train models, or serve them as endpoints across NVIDIA, AMD, and Apple GPUs (coming soon). Gridweave handles scheduling, memory allocation, and multi-GPU scaling so you don't have to.

Get started

Follow the guides in order — each one is a copy-paste path from a blank machine to a working cluster:

  1. Deploy a cluster — master wizard + join GPU nodes.
  2. HTTPS — your own app. / platform. domains with auto-TLS.
  3. Users & credits — signup, onboarding, free vs paid balances.
  4. Get paid with Stripe — users buy credit packs by card; one key, webhook auto-provisions.
  5. Install the SDK & run a job — auth, first GPU job, notebooks.
  6. Update the cluster — ship new versions, keep all state.

Look things up

  • Architecture & operations — every port, expected times, storage layout, teardown.
  • SDK reference — the full Python API: remote(), train(), serve(), gather(), storage.
  • Services (sidebar) — how each internal service works.

The full architecture diagram and feature list live in the project README — this site and the repo are the same markdown, and every page here has an edit pencil that opens the file on GitHub.