GitHub Copilot is an AI developer platform from GitHub that has grown from code completion into a full suite spanning the software development lifecycle, distinguished by multi-model choice that lets developers switch the underlying LLM (such as Anthropic's Claude, Google's Gemini, or OpenAI models) per task. Its capabilities include Copilot Workspace for agentic issue-to-pull-request workflows, Next Edit Suggestions that predict the next edit location, enterprise-wide repository indexing, a Copilot Chat extensions platform, and Copilot Autofix for automated security remediation. It runs across VS Code, Visual Studio, JetBrains IDEs, Xcode, a gh CLI extension, and the GitHub mobile app. It targets individual developers, engineering teams, and enterprise organizations on a subscription model.
GitHub Copilot is GitHub’s AI developer platform, integrated across the software development lifecycle. It began as a pair-programmer for code completion and has grown into a broader suite built around a multi-model approach: developers can choose the underlying LLM to fit the task — fast reasoning, large context, or creative generation.
Multi-model choice
Copilot Chat lets you switch models rather than lock into one — for example Anthropic’s Claude 3.5/3.7 Sonnet for complex coding, Google’s Gemini 1.5 Pro for very large context windows (up to 2M tokens), or OpenAI’s o3-mini for high-speed reasoning.
Copilot Workspace (agentic workflow)
A task-centric environment that takes a GitHub Issue to a Pull Request in one flow: it analyzes the issue, proposes a natural-language plan, generates a spec, and implements the code across multiple files. If the plan is off, edit the natural-language steps and the AI regenerates the implementation to match.
Context and completion
- Next Edit Suggestions (NES): predicts where you are likely to edit next and suggests the change, speeding up refactoring.
- Repository indexing: for Enterprise, Copilot indexes the whole organization’s codebase and answers architecture questions (“How does our authentication service talk to billing?”) with knowledge of proprietary and legacy code.
Chat and extensions
- Extensions platform: Copilot Chat supports third-party extensions, so you can query external tools from the IDE (“Check Sentry for recent errors in this function,” “Deploy this to Azure”).
- Inline chat: highlight code and ask Copilot to refactor, debug, or explain it without leaving the editor.
Copilot Autofix
Integrated with GitHub Advanced Security, it detects vulnerabilities such as SQL injection or XSS and generates a pull request to fix them.
Supported environments
- IDEs: VS Code, Visual Studio (2022/2026), JetBrains (IntelliJ, PyCharm), and Xcode with full Chat support.
- CLI: the
gh copilotextension generates shell commands and explains terminal output. - Mobile: Copilot Chat and agent sessions via the GitHub mobile app.
Where it fits
Legacy modernization through Workspace, incident response via monitoring-tool extensions, test-driven development (write tests first, generate the implementation), and cross-language porting that respects the target language’s idioms.
Direct link: https://github.com/features/copilot
- Multi-Model AI Choice
- Copilot Workspace
- Next Edit Suggestions
- Copilot Autofix
- AI Pair Programming
Ask Lynx Assist
More Guides
Run disciplined SEO A/B tests in seven steps — one metric, two variations, randomized segments, run to significance, track, analyze the winner, and iterate.
Build a topic cluster in seven steps — select and score a pillar, validate it, map subtopics, align to intent, architect internal links, publish, and measure.
Prepare your site for AI search in five steps — content architecture, entity consistency, E-E-A-T, structured data, and machine-readable structure.
Get your content cited by AI in seven steps — answer capsules, link-free extraction, original data, digital PR, community presence, consistent messaging, and tracking.
A seven-step walkthrough for setting up Google Search Console on a new site — property type, DNS verification, sitemap, GA4 link, users, URL checks, and a monitoring routine.