Greptile uses GPT-6 Astra to test running apps, not just read code

Testing applications at scale with GPT-6 Astra

Greptile builds AI agents that review and test code changes with full context of the codebase. "We are in the business of reducing bugs and making the world's software more reliable," says Co-founder and CEO Daksh Gupta, according to OpenAI. For Gupta, that mission requires testing an application's behavior as well as reading its code.

Validating a pull request means checking whether the code is logically correct, does what the developer intended, is secure, and avoids breaking existing functionality. For some of those checks, the application has to be run, because the code on its own may not reveal the problem.

Gupta gives the example of a button that disappears when a website is viewed on an iPhone. With GPT-6 Astra, Greptile can start a development server for a pull request, resize the browser to an iPhone-sized viewport, and simulate a user moving through the application. The model's visual understanding lets it inspect the screen and spot the hidden button.

"A QA firm testing all of your changes is just impractically slow considering how fast we're able to ship code with these models," he says. Greptile can now automate this kind of application testing at scale, according to the source.

Security with Daybreak

Gupta says more powerful models also make it easier for attackers to exploit applications. That puts companies at risk, including some that would previously not have been typical targets.

"We want to make it very easy for even a startup to secure their systems," Gupta says. He adds: "With Daybreak, we're able to do this in a way that is very safe, and we can ensure that our customers have more sophisticated defense mechanisms than the adversaries have offense mechanisms."

Fewer bugs and less rework with Codex

Greptile's Knowledge Base indexes a company's codebase and documents how it works in a form that both people and agents can read. The company makes that documentation available to Codex through a Model Context Protocol (MCP) server. This allows Codex to write code that follows the customer's existing patterns. According to Gupta, thousands of customers have connected Codex to Greptile's knowledge base.

To measure the impact, the team tracks how often AI-generated code is reverted. "When we provide agents with context, fewer reverts happen and customers save time and save inference," Gupta says. Customers can put those savings toward new work instead of reworking the same changes.

Greptile also uses Codex to build its own software and agents. "We are massive users of Codex," Gupta says. "We have enormous amounts of usage across every single function, engineering in particular."

What Ultrafast could change

Gupta explains that the only real bottleneck for Greptile is how quickly the team can turn ideas into working software. He says model intelligence is already sufficient for many tasks, so he sees Ultrafast as a way to speed up development.

His longer-term goal is to make changing and shipping critical enterprise software as easy as shipping a side project, with OpenAI helping Greptile build the validation that requires. For teams shipping code quickly with AI models, the account points to testing and review as the part that needs to keep pace.