October 13, 2026
Startups
Higgsfield brings longer-form video to creators with GPT-6 Astra, achieves precise visuals with GPT-Image-2.5, and paves the way for voice filmmaking with GPT-Live.
Company size: Startup
Region: North America
Industry: Media, Entertainment & Sports
Products: API, Codex
- Making longer videos possible with GPT-6 Astra
- More precise images for small businesses with GPT-Image-2.5
- Enabling a faster path from creative idea to product with Codex
- Looking ahead to voice-to-video with GPT-Live-1
Higgsfield AI is on a mission to empower anyone with a creative idea to produce professional-grade videos and images. Before GPT-6 Astra, these videos were capped at about one minute. Today, CEO and co-founder Alex Mashrabov says users can generate videos over 10 minutes long.
Enabling longer videos is just one way Higgsfield AI is helping small creative teams take on more ambitious projects. Alex notes that with models like GPT-6 Astra and GPT-Image-2.5, it is now possible for a solo creator to execute a high-end fashion campaign entirely from start to finish.
“For Higgsfield, what matters most is seeing small creative teams achieve even greater success powered by models like GPT-6 Astra and GPT-Image-2.5.”
—Alex Mashrabov, CEO and Co-Founder, Higgsfield AI
Making longer videos possible with GPT-6 Astra
Higgsfield AI’s creative agents handle repetitive production tasks, such as generating shot variations. With GPT-6 Astra, these agents can also call external tools to handle the tedious, complex manual work involved in 3D modeling. Alex says this capability helps solve one of the most stubborn challenges in creative AI: maintaining consistency across longer-form videos.
That impact is evident on Higgsfield AI’s own YouTube channel. With nearly 500,000 subscribers, the channel teaches creators how to use AI models and workflows. In the past, producing this content required significant manual effort from Higgsfield AI. But using GPT-6 Astra, the team built a workflow that leverages AI to generate tutorial videos over 10 minutes long.
The ability to produce longer educational videos also holds personal significance for Alex. He has long sponsored rural schools in Kazakhstan and describes an early initiative where they used GPT-6 Astra’s visual understanding to identify concepts students struggle with and tailor instructional videos directly to their learning needs.
More precise images for small businesses with GPT-Image-2.5
For small businesses creating promotional content, legible text and accurate infographics are often just as vital as the imagery itself. Alex notes that GPT-Image-2.5 enables businesses without large production budgets to produce this kind of material.
“GPT-Image-2.5 is state of the art, and it’s the first model of its kind that can reliably render text and infographics on top of images,” he says.
Enabling a faster path from creative idea to product with Codex
Higgsfield AI’s product development relies on tight collaboration between roughly 200 creatives and 100 engineers. Alex shares that Codex empowers creative teams with “the capabilities of strong software engineers.”
Creatives use Codex to turn their ideas into software prototypes. Previously, shipping these prototypes to production required heavy involvement from front-end and back-end engineers. Now, Alex says GPT-6 Astra reduces the effort required to integrate these ideas into the product.
“GPT-6 Astra can deliver production-ready code, which was simply not possible before.”
—Alex Mashrabov, CEO and Co-Founder, Higgsfield AI
With fewer hurdles between prototyping and production, Alex says Higgsfield AI can build and ship better products much faster.
Looking ahead to voice-to-video with GPT-Live-1
As a next step, Higgsfield AI wants creators to be able to simply describe their ideas and watch scenes take shape as they talk. Alex calls this vision “voice to movie”—where AI understands the creator’s spoken intent in real time and renders their vision.
He believes OpenAI’s advancements in planning, tool calling, image generation, and voice are putting this experience within reach, praising the voice model as “state of the art in both speech comprehension and casual conversation.”
For Higgsfield AI, the goal is to make filmmaking feel more like articulating an idea, leaving the heavy lifting in between to AI.