There seems to be an announcement of new AI videos every day, and the latest is from Hedra, a startup that is taking a character-first approach to moving ideas

This week alone, we saw new features in the Luma Labs Dream Machine and the announcement of the new Sora-like Gen-3 on Runway

Character-1 is a research preview of the upcoming foundation video model that gives users granular control over how they animate virtual characters using AI

In the preview, you can give it an audio and a picture Unlike other lip sync tools as it creates a lip sync video showing that character in your photo story, this adds a higher level of expression and movement than I've seen before

Hedra is free to create videos of any length during this research preview The company uses it to test issues with both models and moderation tools before deploying more advanced features

Character-1 is a new underlying AI model designed to create fully controllable and realistic characters using AI The company says they would be able to make expressive stories, songs and even rap in potentially endless periods

For now, it's pretty easy to use it Once registered, you generate audio from the text or give it your own audio and create a character It can generate pictures, AI images, or images in Hedra with text Then just click "Generate video" and wait

There are similarities in features with lip sync tools in some open source projects, research previews, and even platforms like Runway and Synclabs, but what makes Hedra stand out for me is the promise of the future and the representation of the video

The company says of its plans for the future: "This is accessible to everyone so that creators have complete control over emotional dialogue, movement and (yes) the whole world"

Since this is the first phase of the new model, there have been teething problems, especially with overly strict moderation AI, but it has not generated There was no problem with the video that was made

It is currently limited to 30 seconds, so if, like me, you have to do it in two sections, you can upload your own image, which seems to work most effectively with Hedra-generated images

Currently, it only offers square-format video, not widescreen or portrait, and has relatively low resolution But this isn't about creating production-ready content, it's a research preview to showcase the features, and it shows what's actually coming

To put it to the test, I created a short story of an alien invasion This will produce 3 alien and 4 human generals characters from the galactic fleet Compared to human acting, it's as wooden as you'll find in student melodramas — for AI-based lip sync, that's what I've seen before