Hedra Launches 'Omnia' AI for Scene Modeling
The company Hedra has launched Omnia, an AI system designed for full-scene modeling that incorporates motion and emotion. The technology aims to move beyond simple AI-generated "talking heads" to enable more complex, directed performances. The launch was announced alongside a $10,000 challenge to encourage creators to use the new system.
- Hedra was co-founded by CEO Michael Lingelbach and CTO Alex Bergman; Lingelbach was a PhD student at Stanford advised by AI luminary Fei-Fei Li before starting the company. - The company has raised a total of $44 million in funding, with a recent $32 million Series A round led by Andreessen Horowitz's Infrastructure fund. - Omnia is part of Hedra's "Character-3" foundation model, which is described as an "omnimodal" system that integrates text, image, and audio to create character-led videos. - The technology is designed to give creators control over camera movement, body motion, and the background scene from a single image and an audio track, moving beyond static "talking head" videos. - Hedra's earlier models have been used to create over 10 million videos, with some content, like a talking baby podcast, going viral and demonstrating the technology's potential for both consumer and enterprise applications. - The global AI video generator market was valued at over $700 million in 2025 and is projected to grow to over $3.3 billion by 2034, with marketing and advertising being the largest segments. - Competitors in the AI video generation space include well-funded startups like Runway and Synthesia, as well as offerings from major tech companies such as Google's Veo and OpenAI's Sora. - The $10,000 competition is designed to encourage creators to push the limits of the Omnia model and generate content that could go viral.
