Higgsfield Extends AI-Generated Video Length to Over 10 Minutes Using OpenAI Models
Higgsfield has extended its AI video generation from about one minute to more than ten minutes per video. The company credits GPT-6 Astra for longer videos and GPT-Image-2.5 for precise visuals, and says GPT-Live opens a path to voice-driven filmmaking.
Skynet Chance (0%): This is a creative-content application, so it has little bearing on AI control or alignment risk. Longer, more realistic generated video adds only a minor misuse concern around deepfakes and misinformation.
Skynet Date (+0 days): A commercial product built on existing foundation models does not meaningfully change how fast uncontrollable-AI risks could emerge.
AGI Progress (0%): Keeping video coherent over ten-plus minutes shows better long-horizon multimodal generation, a modest capability gain. Most of the advance comes from the underlying OpenAI models rather than Higgsfield's own research.
AGI Date (+0 days): Wider commercial adoption of frontier multimodal models slightly reinforces investment momentum. The effect on the overall AGI timeline is negligible.
<< All AI news for October 13, 2026
[ Get the daily index digest on Telegram → ]