Higgsfield Enabled Ten-Minute AI Video Generation
The company’s GPT-6 Astra tool now extends AI-generated video length beyond the previous one-minute ceiling.
Updated on Oct. 13, 2026 in Artificial Intelligence

Live Poll
Are you optimistic about the future of AI-generated video tools in your daily life?
Higgsfield announced a tool utilizing GPT-6 Astra that allows users to extend AI-generated videos to more than 10 minutes. This development marks a significant expansion from the platform's prior one-minute limit for generated clips.
Why it matters
The shift toward longer-form AI video generation capabilities reflects a maturing production stack aimed at professional marketing and cinematic applications. This capability now serves a reported user base of 32 million people.
The new GPT-6 Astra tool expands video duration to over 10 minutes, a tenfold increase over the previous one-minute limit. This update integrates with GPT-Image-2.5 and provides access to a library of over 50 video and image models via the company's developer API.
The players
Higgsfield
A generative AI startup providing a model API for marketing and cinematic video production that reached a $1 billion annualized revenue run rate.
The details
The tool leverages GPT-6 Astra to maintain temporal coherence over extended sequences, shifting away from the limited temporal windows of previous models. The platform also incorporates GPT-Image-2.5, which functions as a refined spatial synthesis layer to ensure visual precision throughout the longer video output. Users can access these generative capabilities through an API that coordinates more than 50 distinct video and image models.
Timeline
Higgsfield was founded in 2023.
The developer API gained access to 50 models on September 16, 2026.
The company reached an annualized revenue run rate of $1 billion on September 29, 2026.
The GPT-6 Astra video extension tool was announced on October 13, 2026.
The Tech Race
This development moves the industry standard for generative video from short-form clips toward 10-minute production capabilities. It follows the broader trend of expanding temporal coherence in synthetic media generation models.
The 10-minute video capability is currently being rolled out to the platform's 32 million users, including the nearly 1 million paying subscribers. This tool requires integration through the Higgsfield developer API, which allows access to the company's broader set of 50 video and image models.
The takeaway
The leap to 10-minute video generation signals that generative AI is beginning to support professional production workflows rather than just short-form social media content. Users should monitor the rollout of GPT-Live for future capabilities in long-form, voice-commanded filmmaking.
What happens next
Higgsfield is currently developing GPT-Live, a future tool designed to enable the production of full-length films through voice-based commands.
Further reading
For more on the current state of generative video models, visit the Artificial Intelligence section.
Live Poll
Are you optimistic about the future of AI-generated video tools in your daily life?








