Higgsfield says new OpenAI models can extend creator videos past one minute, sharpen visuals and move toward voice filmmaking.

Higgsfield said it can now take AI-generated video from roughly one minute to more than 10 minutes, using OpenAI models including GPT-6 Astra, while also citing GPT-Image-2.5 for more precise visuals and GPT-Live as a step toward voice filmmaking, according to an OpenAI post published on 2026-10-13.

The OpenAI source material is brief and does not provide technical benchmarks, pricing, or rollout details. It frames the development as a creator-focused expansion in video length and visual control, with voice filmmaking described as a future path rather than a finished product. The source does not specify which workflows, customers, or content types are already using the capability.

The announcement follows a period of rapid company expansion reported by Higgsfield in recent weeks. On 2026-09-29, the company said it had crossed a $1 billion annualized revenue run rate and reported more than 32 million users and nearly 1 million paying subscribers in the same announcement. Those figures are company claims, not independently audited numbers. Earlier, on 2026-09-16, Higgsfield said its standalone developer API offered access to more than 50 video and image models. On 2026-08-17, reporting cited a $400 million Series B at a $5.4 billion valuation.

Higgsfield, founded in 2023 by former Snap executive Alex Mashrabov, has positioned itself around creator tools spanning image generation, cinematic video production, marketing workflows and a developer API. The new OpenAI-mentioned capabilities fit that broader product set, but the source material does not establish how much of the longer-video output is available generally versus in limited access or experimental settings.

Source link