Higgsfield Unveils Miracle, a GPT 6.2 Astra Powered Tool That Splits Video Into Layers for Precise Edits
We built Higgsfield Miracle to address one of the biggest challenges in AI video: editing a specific part of a shot without changing the rest. Powered by GPT-6.2 Astra, it splits any video into layers for pixel-level editing in real time, with world knowledge built in. Every shot keeps the image depth that cinema-grade post-production needs. Now in private beta with select testers. Coming soon to Higgsfield MCP. Set up Higgsfield plugin in ChatGPT to be the first to try it.
Context
Editing one part of a shot without changing the rest is a core weakness of AI video, and Miracle is one of the first products built on GPT 6.2 Astra.
Higgsfield is an AI video company that offers creation tools and a plugin inside ChatGPT. OpenAI released GPT 6.2 Astra on Nov 17. Miracle is in private beta with select testers.
Deep dive
Higgsfield posted that it built Higgsfield Miracle to address one of the biggest challenges in AI video: editing a specific part of a shot without changing the rest. The post includes a demo video.
According to the company, Miracle is powered by GPT 6.2 Astra, OpenAI's new frontier model released the same day. Higgsfield says it splits any video into layers for pixel level editing in real time, with world knowledge built in, and that every shot keeps the image depth that cinema grade post production needs.
The tool is not open to the public yet. Higgsfield says Miracle is in private beta with select testers and is coming soon to Higgsfield MCP, its connector for AI assistants. It asks users to set up the Higgsfield plugin in ChatGPT to be among the first to try it.
The launch post drew more than two million views, far more than most product announcements in this deck, though the engagement is modest relative to its reach. The claims about real time layer separation and preserved depth come from Higgsfield's own description and demo; no independent tests were available in the posts reviewed here.
The announcement is an early example of companies building products on GPT 6.2 Astra within hours of its release, and it targets professional video editing rather than generation from scratch.
What to expect next
The main test is how Miracle performs once testers outside the company use it on real footage, especially on fine edits such as faces, hands and moving backgrounds. Watch for a public release date, pricing, and the arrival of the MCP version. Other AI video companies may answer with similar layer based editing, and OpenAI may highlight Miracle as a showcase for what GPT 6.2 Astra can do with video.
