Creating Web Apps has never been so easy: this is how Spark works, the new AI tool from GitHub

GitHub has introduced its new experimental project: GitHub Spark, a tool that allows the creation of small web applications (Web Apps) using natural language. Launched from the GitHub Next labs, this system enables users to give instructions on how to develop apps that can then be refined in several steps.

Subscribe to the Softonic newsletter and get the latest in tech, gaming, entertainment and deals right in your inbox.

Subscribe (it's FREE) ►

Currently in technical preview phase, Spark operates on a GitHub repository and uses GitHub Actions and the Azure CosmosDB database. This way, more experienced developers can view and edit the code if they wish, although the tool is really designed for anyone to use.

To launch an app in Spark, just give an initial request and, within seconds, the user will be able to see a real-time preview. From there, adjustments can be made, and the application can be modified by interacting with the bot. Thomas Dohmke, CEO of GitHub, explained to TechCrunch that Spark does not aim to replace professional developers, but rather aims to offer a tool to explore ideas, create small bots, and help develop software from natural language. “We believe this is the next step in personal software creation,” stated Dohmke.

Although in the GitHub materials Spark is mentioned as a creator of “microapplications,” Dohmke commented that there are no specific limits regarding the complexity of these applications, which leaves room for GitHub users to explore their possibilities. In fact, GitHub also allows apps made with Spark to integrate with any web API and use their own artificial intelligence models, which could further expand their uses.

Dohmke emphasized his excitement to see how far users can go, noting that “it will be very exciting to see how big an application created solely with natural language can be.” Spark also allows choosing between different language models from Anthropic and OpenAI, providing more flexibility to developers.

Users can share their web applications created with Spark with custom permissions. Similarly, if they decide to further develop their projects, they can access the source code to modify it, which is useful when AI makes mistakes.

GitHub DOWNLOAD