Dreamina just relaunched as a full AI film studio 🎬

Good morning. ByteDance relaunched Dreamina as an all-new web experience, an all-in-one AI studio for making video from start to finish. Nvidia is buying Hugging Face, the hub where most open AI models live, for $12.93 billion. And "anotherworldcore" dreamcore videos are all over the feed, those eerie, nostalgic clips of places that feel like a half-remembered dream. Are you team free AI tools or do you pay for the premium ones? Hit reply and let me know! Let's get into it.

ByteDance rolled out an all-new web experience for Dreamina, rebuilding it into an all-in-one AI studio for making video from start to finish. If you haven't used it, Dreamina is ByteDance's AI image and video tool, and this is a full overhaul of the web app, not just a bolt-on feature.

Their idea is doing everything in one place: text, image, video, and audio together, so you're not bouncing between tools to finish one video. It's built around two entry points, Generate for quick prompts and Canvas for hands-on work, plus an AI agent that runs multi-step projects. They're also leaning into teaching, opening up the actual workflows behind breakout short-form videos so you can copy the process instead of just admiring the result.

If you make AI videos and you're tired of learning multiple tools for one clip, it's worth seeing how much of your workflow this collapses into one tab with Dreamina.

Nvidia agreed to buy Hugging Face for $12.93 billion, pending regulatory approval. If the name doesn't ring a bell, Hugging Face is basically the GitHub of AI, the main hub where open models get shared, including a lot of the open image, video, and voice models that smaller tools are quietly built on.

Why it matters, even if you never touch the site, is that Nvidia would now own a huge chunk of the open-model stack on top of already making the chips everything runs on. Nvidia says Hugging Face stays open to everyone and nobody gets forced into its products. The optimistic read is more money behind open models. The nervous read is one company owning more and more of the pipeline.

Day to day, nothing changes for your workflow. But a lot of the free and open tools creators lean on live downstream of Hugging Face, so it's worth knowing who the new owner is. Keep an eye on whether "open" still means open a year from now.

The aesthetic taking over the feed right now is dreamcore, those surreal, quiet clips of places that feel weirdly familiar, a lone phone booth in a field, a vending machine in an empty sunset, a suburban street glowing pink. They run under a trending "anotherworldcore" sound with captions like "am i dreaming?" One of these clips is past 12 million views and the sound has thousands of reels on it.

It works because it hits a specific nerve, nostalgia for a place you've never actually been. The images look like a memory you can't quite place, so people stop, feel something, and tag a friend saying this looks like a dream I had. It's low on action and high on mood, which is exactly what makes it so easy to watch on loop and share.

Here's the playbook. Generate a set of liminal, slightly-off dreamscapes in an image model, keep the palette soft and nostalgic, then animate them into slow 10-second clips with an image-to-video tool. Run them all under the same trending sound, keep the captions short and a little wistful, and post consistently in one visual world so people follow for the vibe. The mood is the brand here, not any single clip.

Stuck on a tool or have an AI video question? Reply, I’ll do my best to help.

Created something cool? Send it over and I might feature you. 📺

📫 Want to get your tool featured or advertise to our readers? Let's chat → [email protected]

Thanks for tuning in!

-Kevin