Free tools Windows power users keep installed
One-click scans. No signup required.
On July 31, 2024, GitHub announced that GPT-4o would power Copilot Chat and pull request summary generation. The announcement described a change at that time—not a promise that GPT-4o remains the model behind every Copilot feature today. GitHub now supports multiple models, and availability depends on a user’s plan and client.
What changed in GitHub Copilot on July 31, 2024?
GitHub said Copilot Chat and pull request summary generation would use GPT-4o. Its official July 31, 2024 Changelog described the change this way: “Copilot Chat and pull request summary generation now use GPT-4o, bringing the performance of OpenAI’s latest flagship model to all developers.”
The announcement covered Copilot Chat in Visual Studio, Visual Studio Code (VS Code), JetBrains IDEs, GitHub Mobile, and GitHub.com. It also gave minimum IDE versions for that update; these are historical announcement-era requirements, not a statement of current compatibility.
| Client | Minimum version listed in GitHub’s 2024 announcement |
|---|---|
| VS Code | GitHub Copilot Chat 0.17+ |
| Visual Studio | 17.8+ |
| JetBrains IDEs | GitHub Copilot 1.5.14+ |
GitHub did not give version minimums for GitHub Mobile or GitHub.com in that announcement. Because client support changes, consult GitHub’s current documentation rather than treating these older minimums as current installation advice.
#1 Best Overall
What performance results did GitHub report?
GitHub’s 2024 announcement reported up to 55% faster time to first byte (TTFT) and a 60% increase in user satisfaction in its testing. These are results reported by GitHub, not independently verified or guaranteed outcomes. “Up to” describes the reported upper bound for TTFT improvement; neither figure predicts how quickly or satisfactorily Copilot will work for a particular person or setup.
Does GPT-4o still power Copilot Chat?
Not as a universal description of Copilot today. GitHub’s current model documentation describes support for multiple AI models and says availability depends on the Copilot plan and client. It lists GPT-4o among utility models, which support background features and cannot be selected or disabled in the model picker.
GitHub’s documentation for Copilot Business and Enterprise base models records GPT-5.3-Codex as the designated base model on March 18, 2026. That information applies to those plans; it does not establish one base model for every Copilot user, client, or feature.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How should you choose a Copilot model?
Where a model picker is available, choose based on the task and the options shown for your plan and client—not on the 2024 GPT-4o announcement alone. GitHub notes that models can have different strengths, and a model choice can affect the quality and relevance of responses. Availability may change.
- Task fit: Consider which model’s strengths suit the work, such as explaining code, drafting a change, or helping investigate a problem.
- Plan and client: Confirm that the model is offered in your account and the Copilot interface you use.
- Latency and accuracy: A faster response is not automatically a more useful or correct one; weigh response time against the quality and relevance you need.
Regardless of model, review Copilot-generated code and validate it—including its security—before using it in production, as GitHub’s model guidance advises.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
