How to use Hugging Face models on VS Code Copilot
- Source
- youtube.com
- Author
- Hugging Face
- Date
Context
Hugging Face engineer Selena explains in a talk that Providers is a single API for models, served by third-party hosts, and says Hugging Face adds no markup. A Hugging Face extension for GitHub Copilot Chat in VS Code builds on VS Code's bring-your-own-key feature, which lets outside providers contribute models to the chat. You add a Hugging Face access , then select models in the model picker.
She shows two choices: fastest, which picks the quickest provider for a model, and cheapest. In a live demo she asked the GLM 4.6 model to implement a GitHub issue for the Hugging Face Python library, and says that kind of request costs roughly 20 to 30 cents. That cost is her estimate for her own demo. She also notes that the extension is open source and ships separately from VS Code releases.
- inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
- open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
- token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
Checking sign-in…
Loading comments…





