OpenAI’s customer story for Perplexity describes how the search-focused company has moved from simply retrieving information to operating real-world systems with GPT-6 Astra. Johnny Ho, Cofounder and Chief Strategy Officer, notes that every improvement in the model’s code-writing ability improves Perplexity‘s search engine as well: it can write better programs that search the web and internal information and summarize it concisely. The harder problem, he says, is applying those informational capabilities to real production systems. With GPT-6 Astra, Perplexity has the model craft communications, edit real-world systems, and monitor production software in ways earlier model generations could not.

The most concrete use case described is testing. Because manual testing time is limited, Ho asks GPT-6 Astra to build a small testing program around an application. The model generates realistic responses similar to what another service would send — for example, a language model API or a connector — and stands in for those services. This lets Perplexity check how the application responds and exercise the workflow from start to finish.

The reported result is that Perplexity now trusts the model with full end-to-end systems and checks in on it much less frequently than with previous generations of models. The story focuses on capability gains and trust, with no metrics, benchmarks, or measured outcomes provided.