YouTube Custom Feeds: Build Your Own Algorithm With AI
YouTube custom feeds are changing how viewers shape what they watch. On Wednesday, the company announced a new feature that lets you describe, in your own words, the kinds of videos you want YouTube to recommend to you. Instead of relying only on the platform’s main recommendation engine, you can now build feeds around your interests, your mood, or your daily routine. The feature was introduced at the Made On YouTube event and will roll out on web and mobile starting next month.
How YouTube Custom Feeds Work
Creating a YouTube custom feed starts with a prompt. You type a description of what you want the feed to be about into a prompt box on YouTube.
The request can be as simple or as detailed as you like. YouTube’s own examples include:
A feed of video podcasts to watch during your 30-minute train commute
A feed of “relaxing commentary videos to unwind with”
A feed built around a specific hobby or interest
Because the feed is created with AI, you can describe at length how you want it curated. You can explain what the feed should feature, what it should exclude, what content to prioritize, and more.
Powered by Gemini AI
The prompt-based feature uses Google’s Gemini AI model to build the feed around your request. Once created, the feed is pinned to the top of your YouTube home page in its own tab.
That placement matters. YouTube says custom feeds will not replace the main feed, which still holds all your standard recommendations. Instead, you can hop straight into whichever feed fits your purpose when you open the YouTube app.
Multiple Feeds and Rollout Timing
Support for creating multiple custom feeds will roll out across YouTube on web and mobile, starting next month. The company announced the timing alongside the feature’s public debut at the Made On YouTube event on Wednesday.
Why YouTube Built Custom Feeds
YouTube frames the feature as a way to help people explore its enormous video library. The platform hosts over 20 billion videos, a scale that can make discovery feel overwhelming.
“One of the amazing things about being on YouTube is that the world is always at your fingertips,” said Emily Moxley, VP of Product Management for Viewer AI, in a pre-brief with reporters ahead of the Made On YouTube event. “There’s over 20 billion videos in the YouTube corpus, so you’re always one search away from a treasure trove of topics.”
Custom feeds aim to turn that endless catalog into something more personal. Rather than searching repeatedly or hoping the algorithm learns your preferences, you can state what you want and get a dedicated space for it.
A Growing Trend Across Platforms
YouTube is not the first platform to hand users more control over their feeds. Several social networks have launched tools that essentially let people build their own algorithms by describing what they want to see.
Decentralized social network Bluesky began popularizing the idea years ago as a way to differentiate itself from rivals like X. This year, Bluesky rolled out an AI tool called Attie that makes feed-building even easier.
Other services have followed with their own feed-building features, many of them AI-powered, including Meta’s Threads, Instagram, X, and, more recently, Spotify.
The pattern suggests a broader shift. Platforms are moving from one-size-fits-all recommendations toward user-defined curation, with AI handling the heavy lifting of interpreting those requests.
What Custom Feeds Mean for Viewers
For everyday viewers, YouTube custom feeds offer a middle ground between the algorithmic main feed and manual search. You still get recommendations, but they follow your stated intent rather than an opaque engagement model.
The feature could be especially useful for recurring needs. A commuter who wants short podcasts every morning can build a feed for that. Someone who wants calm background content in the evening can request it directly. A viewer deep in a niche hobby can ask for videos on that topic without training the main algorithm.
Limits to Keep in Mind
Custom feeds are an addition, not a replacement. The main YouTube feed remains the default destination for recommendations. Custom feeds sit alongside it as pinned tabs, giving you a shortcut when you want something specific.
The feeds are also AI-generated, which means results depend on how clearly you describe your request. Vague prompts may produce feeds that miss the mark, while detailed prompts can specify inclusions, exclusions, and priorities.
YouTube’s move reflects a wider competition over how recommendation systems work. For years, platforms controlled the algorithm entirely. Now, users are being invited to co-author it through natural language.
That shift raises its own questions about transparency and control, but YouTube’s framing is practical: with 20 billion videos available, people need better ways to find what they actually want.
Custom feeds will begin arriving next month on web and mobile. For viewers who have long wanted a say in what YouTube shows them, the prompt box may be the first real opportunity to get it.
