Import from GitHub
Already have a model on GitHub? Import it as a ChipAI.dev model instead of re-creating it by hand. Open Models → Import from GitHub.
It's a three-step wizard:
1. Connect GitHub
Authorize ChipAI.dev to read your repositories. The connection grants:
- Read access to your repositories
- Permission to import model files and the README
- Permission to mirror the repo into ChipAI.dev git storage
You must be signed in to ChipAI.dev before connecting.
2. Pick a repository
Search your repositories and select the one to import. The list shows visibility (public/private), primary language, star count and default branch.
3. Configure the model
ChipAI.dev pre-fills the details from the repo — you can adjust any of them:
| Field | Required | Pre-filled from |
|---|---|---|
| Name | ✓ | Repo name (title-cased) |
| Slug | ✓ | owner-repo, lowercased & URL-safe — permanent |
| Summary | — | Repo description |
| Task | ✓ | Defaults to LLM — pick LLM, Vision, Speech, Multimodal, Edge AI or Robotics |
| License | — | MIT, Apache 2.0, CC BY 4.0, Llama 3, Gemma, GPL 3.0, Other |
| Languages | — | Repo's primary language |
Click Import Model to finish.
What happens on import
- A model record is created.
- The repo's files are mirrored into ChipAI.dev git storage when possible. If mirroring can't complete inline, the files sync on first access instead — either way the model is ready to view.
- You're taken to the new model at
/models/<slug>.
From there, manage it like any other model — add weights via the file browser and edit the card. See Publishing Models for the rest of the workflow.
Next: Using & Downloading Models · Browsing & Discovery · Publishing Models