Pollo AI Launches Video Agent with OpenAI and Aims for 12 Months

Pollo AI introduces a guided video creation assistant powered by OpenAI models, including GPT‑6 Astra, GPT‑Image‑2.5, and GPT‑5.6 for routing. The company claims over 26 million users and states that these tools save more than 50% of the time spent selecting models. Features like Photo to Video Ads also target small brands, allowing for productions in just a few minutes. The company aims to make AI video a daily reflex in the coming year.

Pollo aims for daily adoption of AI video in 12 months

Pollo AI wants AI-generated video to become an integrated component of daily content creation within the next 12 months. According to Bill Zhu, this depends on the ability to easily transform an idea into publishable, testable, or campaign-ready content, which the company links to the continuous improvement of OpenAI models and image generation. Bill Zhu believes that integrating AI into content creation processes will lead to tools that make the process more systematic, with more standardized formats, reduced unsuccessful trials, and streamlined workflows around the models. Pollo AI indicates that users spend over 50% less time selecting or modifying models. For example, one user generated a video ad of about 15 seconds from a simple product photo, while a jewelry brand produced an editorial image featuring its creations worn by a model sitting next to a cheetah using Pollo Agent.

A guided agent built on GPT‑6 Astra, GPT‑Image‑2.5, and GPT‑5.6

Pollo AI's latest offering, Pollo Agent, provides a fully guided journey for video creation, relying on OpenAI models like GPT‑6 Astra and GPT‑Image‑2.5. Pollo Agent supports the interpretation of initial ideas and visual references provided by the user, then constructs scenarios, scene sequences, and scripts. To balance reasoning depth, execution speed, and cost, Pollo AI employs GPT‑5.6 for routing and uses GPT‑6 Astra for more advanced tasks, such as creating a narrative framework or planning scene modifications. This combination of models allows for adjusting the level of reasoning based on user demand. Peter Zhou, CTO, highlights the ability of OpenAI models to handle ambiguous creative instructions as well as their multimodal versatility, while emphasizing the positive results achieved with GPT‑5.6. Bill Zhu notes that improvements in the models have made more workflows feasible, with simpler inputs and more reliable outputs.

Claimed usage: 26 million users and 30% of video sessions

Pollo AI claims over 26 million users who create videos, ads, and visual content on its platform. The team's initial breakthrough focused on video models, and today, 30% of creator sessions begin with these models or trending format workflows. The majority of users are content creators, with a particular popularity for micro-dramas featuring human actors and AIs. Pollo AI is also seeing growing interest from marketers and business owners looking to quickly produce video content. The company reports a gain of over 50% in time spent selecting or changing models.

Marketing tools to produce in minutes rather than days

The Photo to Video Ads tool, part of Pollo AI's Marketing Studio suite, allows users to convert a static image accompanied by a text prompt into a short video using OpenAI models. It is designed to give small businesses and DTC brands the ability to create product ads for social media in just a few minutes, compared to several days previously. For example, one user generated a perfume ad from a simple product photo, resulting in a video of about 15 seconds without needing a filming crew. A jewelry brand also utilized Pollo Agent and GPT‑Image‑2.5 to bring to life a creative idea featuring its jewelry worn by a model sitting next to a cheetah.

Positioning and target: simplifying video beyond experts

Pollo AI was founded to enable anyone to transform creative ideas into finished images and videos without production expertise, in a context where the diversity of tools and models can discourage some creators and marketers. Bill Zhu, founder and CEO, summarizes the company's ambition: to make video creation as accessible as design has become through tools like Canva, in order to broaden usage beyond experienced professionals. Peter Zhou serves as CTO at Pollo AI.