Pollo AI was founded to let people transform creative ideas into finished images and short videos without production expertise. The company says that even with generative AI, creators and marketers often struggle to pick tools and models, which can stall projects — a gap Pollo AI aims to close by simplifying workflows.
Users and usage patterns
More than 26 million people use Pollo AI to create videos, ads and visual content. The company’s early breakthrough came from video templates: about 30% of creator sessions now start from templates or trending-format workflows. Building on that, Pollo AI launched Pollo Agent to offer a more guided creation experience.
Pollo Agent and OpenAI models
Pollo Agent interprets rough concepts and visual references from users, then generates storylines, scene flows and scripts. The product uses OpenAI models including GPT‑5.6, GPT‑6 Astra and GPT‑Image‑2.5. Pollo AI CTO Peter Zhou said OpenAI’s models stood out for understanding ambiguous creative inputs and for multimodal capabilities; the team specifically noted that “GPT‑5.6 performed especially well there.”
To match reasoning depth and cost to task complexity, Pollo AI routes simpler tasks to GPT‑5.6 and assigns harder work — like narrative development and scene revision planning — to GPT‑6 Astra. According to the company, this model mix reduces the time users spend choosing or switching models by more than 50%, freeing up time for creative work.
Image generation and campaign examples
Pollo AI evaluates generative models against real user prompts and workflows, assessing output quality, consistency, stylistic range, speed and cost. For image generation they selected GPT‑Image‑2.5 to produce posters, storyboards and campaign visuals.
The platform’s main users are content creators, particularly those making short micro‑dramas that mix AI and human actors, but marketers and small business owners are increasingly using Pollo AI to produce video content quickly.
Examples the company cites:
-
Photo to Video Ads: Part of Pollo’s Marketing Studio, this tool converts a still photo plus a prompt into a short video. One user turned a product photo into roughly a 15‑second fragrance ad; the tool generated velvet textures, warm spotlights and bottle close‑ups, finishing with the perfume framed against deep red fabric.
-
Editorial campaign for an jewelry brand: With no budget for a high‑concept studio shoot, a brand used Pollo Agent and GPT‑Image‑2.5 to realize a vision of statement jewelry on a model seated beside a cheetah. The generated image used dark drapery and warm lighting to keep the jewelry prominent, producing a luxury campaign visual with lower production effort and cost than a live shoot.
-
Playful beverage campaign: A beverage brand wanted a whimsical feel for a peach soda. Using Pollo Agent and GPT‑Image‑2.5, the team transformed a simple bottle photo into a snowy miniature landscape complete with a train, lizards and sparkling droplets, resulting in a creative image for campaign use.
GPT‑Image‑2.5 was chosen for its sharper details and faster generation in creative workflows, according to Pollo AI.
Goals and outlook
Pollo AI aims to make AI video a routine part of content creation over the next 12 months. Founder and CEO Bill Zhu says the key is making it as easy as possible to go from an idea to an asset ready to post, test or include in a campaign — an outcome the company plans to deliver as OpenAI’s models and image generation continue to improve.
Zhu argues the AI shift will favor products that make creation repeatable from the outset, with stronger formats, less trial and error, and simpler workflows around the models themselves.
Technical highlights
- Models used: GPT‑5.6 (routing and control), GPT‑6 Astra (narrative and scene planning), GPT‑Image‑2.5 (image generation).
- Users: more than 26 million.
- Sessions starting from templates: ~30%.
- Reduced time spent selecting models: over 50%.
- Time horizon: aims to normalize AI video in everyday creation within 12 months.
By combining OpenAI models and template‑driven workflows, Pollo AI automates the conversion of creative briefs and photos into finished images and short video ads while emphasizing model stability and release cadence as development priorities.
