Higgsfield uses GPT to extend AI video generation
What happened
On October 13, OpenAI News reported that Higgsfield used GPT-6 Astra to push AI-generated video from about 1 minute to more than 10 minutes, and used GPT-Image-2.5 to generate readable text and infographics. The report also said its roughly 200 creatives and 100 engineers turned ideas into software prototypes with Codex, and that GPT-6 Astra can output production-ready code. Higgsfield's next step is to use GPT-Live-1 for "voice to film."
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
- OpenAI NewsHiggsfield takes AI video from 1 minute to over 10
Higgsfield 借助 GPT-6 Astra 将 AI 生成视频时长从约 1 分钟提升至 10 分钟以上，并用 GPT-Image-2.5 生成可读文字与信息图。其约 200 名创意人员与 100 名工程师通过 Codex 将创意转为软件原型，GPT-6 Astra 还能输出可上线的生产级代码。Higgsfield 下一步计划用 GPT-Live-1 实现"语音转电影"。
Heat over time
Not enough continuous observations to draw a trend yet.