Pollo AI builds its video agent on OpenAI's GPT-6 Astra and GPT-Image-2.5

A guided workflow for people without production skills

Pollo AI was founded on the idea that people should be able to turn creative ideas into finished images and videos without production expertise. The company notes that even with generative AI, creators and marketers can struggle to navigate different tools and choose between models, and some abandon their projects out of frustration.

"We want to do for video what Canva did for design," said Bill Zhu, founder and CEO of Pollo AI. Today the platform is used to create videos, ads and visual content. Its first breakthrough was video templates: 30% of creator sessions now start from these templates or trending-format workflows.

The latest product is Pollo Agent, which uses OpenAI models, including GPT-6 Astra and GPT-Image-2.5, to provide a fully guided video creation workflow. It interprets a user's rough ideas and visual references, then develops storylines, scene flows and scripts.

How the model mix works

CTO Peter Zhou said OpenAI models stood out for understanding ambiguous creative inputs and for their multimodal capability. "GPT-5.6 performed especially well there," he said. The team also values the stability and release cadence of the OpenAI API.

To balance reasoning depth, speed and cost, Pollo AI uses GPT-5.6 for routing and GPT-6 Astra for harder tasks, such as developing a narrative or planning scene revisions. This lets the system match the level of reasoning to the user's prompt. Pollo AI reports that users spend more than 50% less time choosing or switching models.

The company tests generative models against real user prompts and workflows, judging output quality, consistency, stylistic range, speed and cost. GPT-Image-2.5 handles image generation for posters, storyboards and campaign visuals. Zhou said: "If we receive any task from a user to generate an image, we always use GPT-Image-2.5."

From product photos to ads and campaign images

Pollo AI's primary users are content creators, and the platform is especially popular for micro dramas with both AI and human actors. More marketers and business owners are also starting to use it. The Photo to Video Ads tool in its Marketing Studio suite turns a still image and a prompt into a short video, aimed at helping small businesses and DTC brands make social media product ads in minutes rather than days.

The source describes three customer examples:

• Fragrance ad: a user without the budget for a professional shoot turned a product photo into a roughly 15-second video with velvet textures, warm spotlights and close-ups of the bottle, ending with the fragrance framed against deep red fabric.
• Jewelry campaign: a brand used Pollo Agent and GPT-Image-2.5 to create an editorial image of its statement jewelry on a model seated beside a cheetah, with dark drapery and warm lighting. According to the company, this took lower production effort, time and cost than a live shoot.
• Peach soda visual: a beverage brand turned a simple bottle image into a snowy miniature landscape with a train, lizards and sparkling droplets for its campaign.

What it means for creators and marketers

Pollo wants AI video to become part of everyday content creation over the next 12 months. Zhu said this depends on making it even easier to turn an idea into an asset ready to post, test or use in a campaign. He said that as OpenAI's models have improved, more of the workflows Pollo wanted to support have become practical inside the product, so users can start from simpler inputs and reach stronger outputs more reliably.

Zhu expects the shift in content creation to favor products that make creation "more repeatable from the start, with stronger formats, less trial and error, and a simpler workflow around the models themselves." For small businesses and brands with limited budgets, the examples suggest a route to campaign visuals and video ads without a studio shoot, though the results described come from Pollo AI and OpenAI themselves.