News · Meta launches the Vibes feed in Europe, wrapping AI video generation in a social frontend
Meta launches the Vibes feed in Europe, wrapping AI video generation in a social frontend
Meta's new AI app puts a remixable video feed, a chat assistant, and glasses management behind one surface — and the feed, not the prompt box, is the organizing interface.
The feed becomes the primary interface for generation
Most consumer AI products still open onto a text box. Meta's new app opens onto a feed. Vibes is described as "a dedicated feed, displaying AI-generated videos and content from creators and communities, which will become more personalised to your interests over time." That single design decision is the story here.
By making a scrollable surface the entry point rather than an empty prompt, Meta reframes generation as something you encounter and modify, not something you originate from scratch. The source is explicit that you "can create your own videos using imaginative prompts, or remix what's already been shared" — and remixing existing content lowers the blank-page cost that makes text-first AI tools intimidating to casual users.
This is a familiar Meta pattern applied to a new medium. The company is treating AI video the way it treated photos and Reels: as feed objects that can be viewed, shared, and re-derived. The frontend does the work of teaching people what the model can do by showing them what others already made.
One app, three jobs, no obvious seam
The app bundles three distinct experiences into one shell: the Vibes feed, a conversational assistant, and management for Meta's AI glasses. The assistant side carries text prompting, image generation and animation, and image editing. The glasses side handles importing, editing, and sharing captured media.
These are genuinely different interaction models — a feed you scroll, a chat you converse with, and a device manager you configure. Stitching them into a single app is a hard frontend problem, because each has its own natural layout and each competes for the home surface. Meta's framing puts Vibes and "creative media sharing" "at the heart of your experience," which suggests the feed wins the default view and the other jobs live one layer down.
Worth noting: Meta AI already exists inside Facebook, Instagram, Messenger, and WhatsApp. The standalone app is a deliberate second front — a place where the assistant and the generative feed are the point rather than a helper bolted onto messaging. That creates a duplication question the announcement does not resolve: which surface a European user reaches for when the same assistant lives in five places.
The distribution loop runs through Stories and Reels
The clearest lever in the announcement is cross-posting. Content made in Vibes "can be shared and posted directly to the Vibes feed, sent to friends, or cross-posted to Instagram and Facebook Stories and Reels." That connects a new, sparsely populated feed to Meta's existing billions-scale surfaces.
For a fresh feed, the cold-start problem is everything — an empty surface has nothing to remix. Routing output into Instagram and Facebook gives Vibes content immediate reach it could not earn on its own, and pulls viewers back toward the app to make their own. The frontend design encodes a growth loop, not just a viewer.
Media generation in the app itself has jumped more than tenfold since it first launched.Montana Labs
That figure, from the US launch, is the one metric Meta offers for whether the feed-first approach changes behavior. It measures generation volume inside the app, not retention or output quality — so it tells us people are prompting more, not yet that they are staying or that what they make travels well.
What a feed-first generator means to build against
The specific implication of this launch is that Meta is betting the interface, not the model, is what unlocks casual AI video creation at scale. It cites over 20 billion images made with its AI tools and frames Vibes as the next surface for that behavior in video form. The wager is that remix-and-share mechanics convert passive viewers into creators faster than a better prompt box would.
For teams building generative features, the takeaway is concrete: seed the surface with existing content, make remixing cheaper than originating, and wire outputs into whatever distribution you already own. Meta can do all three because it controls Reels and Stories. The harder question for everyone without that reach is whether a feed-first generator works at all without a pre-existing audience to fill it — and this launch will test that in a new European market where the feed starts near empty.
Find this story relevant to you?
Contact us to find a unique solution
Need an AI engineering partner that can actually build?
We help businesses integrate AI, build AI-powered products, automate high-value workflows, and modernize the software systems behind them.
Related reading
More analysis around product delivery, operational AI, and the systems work that makes deployment hold up in reality.