Airbnb has expanded its agreement with OpenAI, giving engineering and product teams broader access to frontier models including GPT-6 Astra through OpenAI APIs and Amazon Bedrock. The move comes as the company reports its development teams are shipping roughly 80% more features than a year ago, with OpenAI models playing a central role in that acceleration.
From code generation to system design
Airbnb engineers already use an internal AI assistant powered by Codex and models like GPT-5.6 Sol, Terra, and Luna for writing software and creating remote AI agents. The new agreement widens access to the latest frontier models. Early use of GPT-6 Astra has shown results beyond coding. The model helps engineers track down hard bugs, shape system designs, and brainstorm engineering approaches. In one test involving strategic documents and other non-coding work, a user reached the desired output in 3-4 passes with Astra, compared with 20 or more rounds using other models.
"Our development teams are shipping roughly 80% more features than a year ago. OpenAI's frontier models, including GPT-6 Astra, are a key element of our developer tooling that helps us keep that momentum going and build better products faster," said Ahmad Al-Dahle, Chief Technology Officer at Airbnb.
AI across the marketplace
Airbnb's use of OpenAI models extends well beyond engineering. The company applies AI and machine learning across search, fraud prevention, guest and host support, and insurance claims. These systems help the company understand what guests are looking for, identify potentially fraudulent content or claims, assist guests when travel plans change, and speed up insurance payouts for hosts.
As Airbnb brings more of the trip into one app-homes, services, experiences, airport pickups, car rentals-the expanded model access gives teams building those features more powerful tools to move faster. "Airbnb has been using AI to rethink how software gets built, and expanding access to models like GPT-6 Astra gives its teams even more powerful tools to move from ideas to production faster," said Dali Rajic, Chief Revenue Officer at OpenAI.
Why this matters for product development and operations leaders
Airbnb's 80% increase in feature output is a measurable benchmark for what happens when development teams get broad access to frontier models. The jump from 20-plus iterations down to 3-4 passes on complex work signals a shift in how fast teams can move from concept to production. For leaders in product, engineering, and operations, the takeaway is concrete: the gap between early AI adopters and those waiting on the sidelines is widening, and it is showing up in shipping velocity, not just internal experiments. Professionals looking to build similar capabilities in their own teams can explore structured learning paths like AI for Software Developers Courses to close that gap systematically.
Your membership also unlocks:
