GitHub has broadened its AI-powered coding assistant, GitHub Copilot, to support multiple generative AI models, allowing developers to select from leading models like Google Gemini 1.5 Pro, OpenAI GPT-4o, and Anthropic Claude 3.5 Sonnet. Initially available in Copilot Chat, this multi-model support enables developers to tailor AI model selection to specific tasks while also providing the option to use Copilot’s default model. Enhanced functionalities also extend to Visual Studio Code, where GitHub Copilot now supports multi-file editing using natural language prompts and Autofix’s security capabilities, enabling large-scale vulnerability remediation.
GitHub additionally introduced GitHub Spark, an AI-native tool designed to facilitate web app development through natural language prompts. With Spark, developers of all levels can create and deploy micro apps, or “sparks,” on desktop and mobile platforms without managing cloud resources. Users can iterate and preview their apps in real-time with support from OpenAI and Anthropic models. GitHub also revealed upcoming features, including GitHub Copilot Extensions, which allow developers to interact with tools like Docker and Stack Overflow. This extension feature will be generally available in early 2025, highlighting GitHub’s ongoing commitment to enhancing developer productivity through integrated AI solutions.