At the GitHub Universe 2024 conference, GitHub announced that its auto-completion and programming tool Copilot will support multiple artificial intelligence models. Developers will now be able to choose models from Anthropic, Google, and OpenAI to work with GitHub Copilot.
Users of GitHub Copilot in the web version or VS Code will be able to choose the Claude 3.5 model, and in the coming weeks, the Gemini 1.5 Pro model will also be available. In addition, the GPT-4o, o1-preview, and o1-mini models from OpenAI will soon be available in GitHub Copilot. When working with Copilot Chat, developers will be able to switch between models to find the best one for a specific task.
Thomas Domke, CEO of GitHub, emphasizes:
«No model can be universal for all scenarios. Developers strive to create projects with models that best suit their needs».
The company also presented Spark, a tool for building web applications using natural language. Spark uses the OpenAI and Anthropic models to generate real-time previews of applications. Experienced developers can edit the code directly, while beginners can build apps exclusively through text commands. Once an app is created, it can be run on a computer, tablet, or smartphone, and shared with others so that people can remix and build new Spark-based apps.
GitHub has announced additional updates to Copilot: starting November 1, the Multi-file edit feature in VS Code will be available for simultaneous editing of multiple files, Copilot Extensions will be available in early 2025, and Copilot for Xcode will be available for public testing.
Source: The Verge
Spelling error report
The following text will be sent to our editors: