What changed
OpenAI’s customer story says Higgsfield moved from producing AI videos around one minute long to sequences longer than ten minutes. The company credits GPT-6 Astra with helping it coordinate a more complex production workflow, including story structure, generated scenes and voice. The headline is useful because duration is an understandable stress test: longer videos require continuity across many more shots than a short prompt-to-clip demo.
This is a customer case study, not an independent benchmark or a general model capability guarantee. OpenAI and Higgsfield do not claim that every user can generate a finished ten-minute film in one prompt. The reported result describes a production process built by Higgsfield, where tools and human direction help assemble a longer piece.
What the result means
For creators, the more interesting question is whether the workflow keeps characters, settings and audio coherent across edits. A longer export matters only if it reduces stitching work without turning revisions into a second production. The source does not publish a standardized quality comparison, so treat the length as a company-reported achievement and judge the actual sample.
Source published 2026-10-13. Coverage is based on the maker’s announcement and demonstration.
