Introducing Playground v3

Our newest AI model Playground v3 (or "PGv3") is purpose-built for graphic design capabilities.

We focused on being the best in the world at prompt understanding and control – pushing beyond aesthetics, which has become saturated as a benchmark. PGv3 outperforms all the most popular image models made in its class, as show in our technical report.

Creating an AI graphic designer also requires introducing new model capabilities. Our research focused on improving control over visual communication, typography, color precision, and layout. We also introduce Argus, a new vision language model that can describe images better than a human could (or would want to). VLMs are a key component to improving our image model so that people to control every element of a design.

To learn more about PGv3's capabilities and performance, go to our PGv3 landing page. And for more specifics into our research methods, evaluations, and benchmark testing, please see our technical report on Arxiv.

We're also hiring: playground.com/jobs