Setting up GitHub Copilot for your development workflow

Explore Copilot’s functionality in online and offline modes

GitHub Copilot is a cloud-powered AI tool that depends heavily on access to online resources for its real-time suggestions. However, developers today work in a mix of environments—some entirely cloud based (like GitHub Codespaces) and others locally installed and sometimes even offline. Understanding how Copilot behaves in both scenarios is essential to plan your workflow, avoid surprises, and maximize efficiency.

GitHub Codespaces integration

GitHub Codespaces is a cloud-hosted development environment that runs entirely in the browser or from Visual Studio Code. It allows you to spin up container-based dev environments with all your tools, dependencies, and extensions—on demand. Copilot integrates natively and seamlessly into GitHub Codespaces, making it ideal for developers who want to:

• Start coding instantly without local setup.

• Work from anywhere with consistent environments.

• Get real-time Copilot suggestions with minimal latency.

Use case example

A student working on a group project can open Codespaces directly from the GitHub repo, write a comment like:

// function to calculate prime numbers in JavaScript

and get an instant, intelligent suggestion from Copilot without any local setup. All suggestions are served from GitHub’s cloud infrastructure.

How Copilot works without an internet connection (limitations)

GitHub Copilot requires an active internet connection for generating suggestions. This is because Copilot doesn’t run its model locally; it sends your prompt (e.g., comments or code context) to the GitHub Copilot servers, which return relevant completions. However, it is available in offline mode, which can be useful if you temporarily lose connectivity (such as on a plane or in a restricted network environment) and still need to keep working with your existing code.

These are the key limitations in offline mode:

• No suggestions: If your machine goes offline, Copilot cannot provide new completions or suggestions. The Copilot icon typically shows a warning or “offline” message in such cases.

• No updates or learning: Copilot cannot fetch new context or respond to changes in real time.

• Cached data is limited: Even if you had previous suggestions, they are not stored locally in a meaningful way for reuse.

• Feature access is blocked: Features like Labs, feedback submission, and telemetry are inaccessible offline.

Use case caution

If you're coding on a plane or in a remote area with no internet, Copilot won’t function. Developers in such scenarios should prepare code templates, snippets, or reference materials in advance, or they can temporarily rely on traditional IDE features like IntelliSense.

GitHub Copilot mode comparison

The following table compares GitHub Copilot’s availability and performance across different development environments, both online and offline.