Use Github Copilot as an llm provider in AI Agent

The idea is:

To allow a user to use GitHub Copilot subscription as a LLM provider in the AI Nodes.

My use case:

Any LLM usage, with Github Copilot as the LLM Provider.

I think it would be beneficial to add this because:

People could utilize their Github Copilot subscriptions as a LLM Providers.

Any resources to support this?

Github Copilot

Are you willing to work on this?

Yes, if anyone bootstraps this.

Is there any possibility that the one could use Github Copilot subscription as a LLM provider?

I really wish this would be possible natively in n8n.

Similar request here : Please add github copilot model support in n8n

Community node but lacking traction here : https://www.npmjs.com/package/n8n-nodes-github-copilot-agent-tool

welcome to the n8n community @antonkristensen :tada:
I was actually looking into this earlier, and the biggest hurdle seems to be GitHub’s Terms of Service and the lack of a standard public API for personal Copilot subs. Right now, they really pin it down to IDEs and their own CLI. Have you looked into whether their current API even allows for this kind of external integration without getting flagged?

1 Like

Hello @jibd @antonkristensen

I’ve just added a new node that integrates the Copilot subscription into your workflows using the Copilot official SDK (still in preview). This SDK will internally spawn a CLI sub-process or remotely connect to one.

I’m rolling it out in beta and would love your feedback. https://www.npmjs.com/package/n8n-nodes-copilot-agent

Two auth options: GitHub PAT or remote CLI, whatever fits your setup.

Spotted a bug or have ideas? Drop them on GitHub. GitHub - yashodhah/n8n-nodes-copilot-agent · GitHub

Hi @tamy.santos,

From what I’ve seen so far (I looked into how opencode and VS Code Copilot Chat integrate with the Copilot models), VS Code Copilot Chat offers a flexible multi-format SDK abstraction that can communicate with OpenAI-compatible, OpenAI Responses or Anthropic Messages protocols… chosen at runtime based on the selected model. All requests pass through GitHub’s managed CAPI ( Copilot API) gateway for authentication quota tracking and backend routing.

1 Like