While Github has been slowly adding AI powered features for developers in the last couple of years, the Microsoft owned company has gone as step ahead with its latest feature called Github Spark that allows users to create an app by just giving prompts in natural language.
The new feature which is an experiment and is part of the GitHub Next labs, gives users the choice between using an OpenAI GPT model or a Claude Sonnet model. Notably, while OpenAI has tried to tune its latest models for developers, Claude’s Sonnet models still tend to make more of a buzz in the developer circles due to technical reasoning and debugging abilities.
With the new tool, GitHub allows users to quickly build a small web app or ‘micro app’ using natural language. Unlike othe GitHub tools, Spark doesn’t just generate code for the required app but instead runs and displays an interactive preview which can then be altered via further interactions.
Interestingly, Spark can also be used to brainstorm ideas using its revision variants option. So for instance if a user isn’t entirely user of how their micro app should look like they can ask the AI to create a set of 3-6 variants of the app, each with a subtle yet meaningful deviation to get better ideas.
Github also provides a number of other options with Spark in order to improve the app building process like single click deployment, Copilot agent collaboration and model selector. Users can choose between Claude 3.5 Sonnet, OpenAI’s o1-preview, o-1 mini and GPT-4o.
Github CEO Thomas Dohke while praising the new feature in a post on X (formerly Twitter) wrote, “In the last five decades of software development, producing software required manually converting human language into programming language, compiling it, debugging it, testing it – and then going back to more coding. Today, we take a step toward the ideal magic of creation: the idea in your head becomes reality in a matter of minutes.”