Higgsfield takes AI video from 1 minute to over 10
Higgsfield has introduced the ability to generate AI videos exceeding 10 minutes in duration. The release includes the integration of GPT-6 Astra, GPT-Image-2.5, and the GPT-Live voice filmmaking tool.
Verified State Diff
Impact & Verification Analysis
Content creators, filmmakers, and digital media production studios.
The transition from short-form to long-form video generation fundamentally changes the utility of AI models from social media snippets to full-scale cinematic production, significantly increasing the competitive pressure on existing video generation platforms.
Full Fact Overview
This announcement marks a significant shift in generative video capabilities, moving from short-form clips to long-form content generation. By integrating GPT-6 Astra for extended video synthesis, GPT-Image-2.5 for visual fidelity, and GPT-Live for real-time voice synchronization, Higgsfield is positioning itself to compete in the long-form cinematic AI production space. This architectural leap suggests a move toward temporal consistency over extended durations, which has historically been a primary bottleneck in AI video models.