Pollo AI's Pollo Agent taps OpenAI for video creation

Pollo AI started with a simple premise: creators and marketers should be able to turn ideas into finished images and videos without production expertise. According to the article, the team's first breakthrough was video templates, and 30% of creator sessions now start from those templates or trending-format workflows. That groundwork set up Pollo Agent, which interprets rough ideas and visual references and then develops storylines, scene flows, and scripts.

The article describes a model mix meant to balance reasoning depth, speed, and cost — GPT‑5.6 handles routing, while GPT‑6 Astra takes on harder work such as developing a narrative or planning scene revisions. Peter Zhou, Pollo AI's CTO, says GPT‑Image‑2.5 always handles image requests because it is especially good with different types of characters and letters. Three customer examples show the range: a roughly 15-second fragrance ad from a single product photo, a jewelry campaign image placed beside a cheetah, and a peach soda bottle turned into a snowy miniature landscape with a train and lizards.

Pollo AI's primary users are content creators, and the platform is described as especially popular for micro dramas using both AI and human actors. The article notes that more marketers and business owners are starting to use it as well. Bill Zhu frames the goal around repeatable formats and less trial and error, and says Pollo wants AI video to become part of everyday content creation over the next 12 months.