Key takeaways
- Hype is a second button under a video that viewers spend deliberately: three a week, each one applied to a long-form video from a Partner Program channel between 500 and 500,000 subscribers, inside that video's first seven days.
- The handicap is real but it is not published as a formula. YouTube documents exactly two points on the curve — 7,500 points per hype at 500 subscribers, 50 at 500,000 — and a power law fitted through those two anchors reproduces both exactly.
- That curve compresses a size advantage rather than inverting it. A channel a thousand times your size, converting the same share of its audience, still scores roughly seven times your points. The variable you control is the conversion rate, not the subscriber count.
- The prize is a browse surface unlike any other on YouTube: up to 50 videos, one country, not personalised. Everyone in your country sees the same fifty tiles in the same order.
- Which makes the leaderboard the one place where you can know your competition before you design against it — and where nobody scrolling past has any idea who you are.
- Hype pays nothing, names nobody, and has never been described by YouTube as a ranking signal. Treat "hype boosts the algorithm" as an unsourced claim.
Almost every distribution mechanism on YouTube works the same way: a viewer does something for their own reasons — watches, skips, clicks, stays — and the system reads that behaviour as evidence. You never ask for it directly. Asking would corrupt the signal, which is precisely why the platform has spent two decades ignoring what creators say to camera and watching what viewers do instead.
Hype is the exception. It is a button a person presses on purpose, on your behalf, knowing it helps you and does nothing for them. It has a weekly allowance, so pressing it for you means not pressing it for someone else. And it leads somewhere concrete: a ranked list of at most fifty videos, one per country, shown identically to everyone who opens it.
Almost nothing has been written about how that list actually behaves. The guides explain the button and stop. So here is the whole mechanism: the eligibility rules including the one everybody gets wrong, the points curve worked out from the only two numbers YouTube publishes, what those numbers imply about who can realistically top a leaderboard, and what happens to your thumbnail when it lands in a fifty-tile grid full of strangers.
What Hype is, mechanically
Hype was announced on 18 September 2024 at Made on YouTube, alongside the first wave of what became the Communities and AI features. It ran as a beta in Brazil, Turkey and Taiwan through late 2024, expanded through 2025, and reached 39 countries on 26 August 2025, including the United States, the United Kingdom, Japan, Korea, India and Indonesia.
On the watch page in the YouTube mobile app, eligible videos carry a Hype button in the action row beneath the player, alongside Like and Thanks. A viewer taps it. The video receives points. Points accumulate into a ranking, and the top of that ranking appears under Explore as a leaderboard. That is the entire loop.
Three constraints shape everything downstream. Each viewer gets three hypes per week, refreshed weekly — they can stack all three on one video or spread them across three. Only long-form videos uploaded within the last seven days can be hyped. And the button exists only inside the YouTube app; neither desktop web nor a mobile browser shows it, though YouTube signalled in November 2025 that a desktop version was coming.
Who is eligible — and the requirement most guides skip
Every article about Hype says "channels between 500 and 500,000 subscribers". That is half the rule. YouTube's eligibility documentation requires membership of the YouTube Partner Program as well, and the 500 figure is not an arbitrary floor — it is the subscriber count at which the Partner Program's fan-funding tier opens.
That tier is not free either. As covered in our breakdown of the 2027 monetisation rules, it asks for 500 subscribers plus 3,000 valid public watch hours in the past year or 3 million valid public Shorts views in the past 90 days. So a channel with 600 subscribers and 400 watch hours has no Hype button, and no amount of asking its audience will produce one. The real gate is the Partner Program, and the subscriber range simply describes who inside it qualifies.
| Condition | Rule |
|---|---|
| Channel | In the YouTube Partner Program, 500 to 500,000 subscribers, in an available country |
| Content | Long-form videos only |
| Window | Uploaded within the last 7 days |
| Not eligible | Shorts; live streams while live; Premieres while live |
| Viewer allowance | 3 hypes per week, stackable or split |
| Surface | YouTube mobile app only — not desktop web, not mobile browser |
The Shorts exclusion is the strategically interesting one. A channel that has built its audience on the swipe feed has, by construction, an audience that never sees a Hype button, because the button does not appear on the format they watch. Hype rewards long-form specifically, and it rewards it in the seven days when a long-form video is newest.
The points curve, worked out
Here is where the published information runs out and the arithmetic starts. YouTube says smaller channels receive more points per hype, and gives exactly two worked examples in its Help Centre: a channel with 500 subscribers earns 7,500 points per hype, and a channel with 500,000 subscribers earns 50. No formula, no table, nothing in between.
Two points define a power law. Across a thousandfold range of subscribers the points fall by a factor of 150, which fits an exponent of about 0.73 — points scale roughly as the subscriber count raised to the power of minus three-quarters. Fit that curve through both published anchors and it reproduces 7,500 and 50 exactly. Everything between them is interpolation, and should be read as an estimate of the shape rather than a leaked table.
| Subscribers | Points per hype | Hypes for 1,000,000 points |
|---|---|---|
| 500 | 7,500 (published) | 133 |
| 1,000 | ~4,500 | ~220 |
| 2,000 | ~2,700 | ~365 |
| 5,000 | ~1,400 | ~710 |
| 10,000 | ~850 | ~1,170 |
| 25,000 | ~440 | ~2,280 |
| 50,000 | ~265 | ~3,760 |
| 100,000 | ~160 | ~6,220 |
| 250,000 | ~83 | ~12,100 |
| 500,000 | 50 (published) | 20,000 |
Read this table as a shape, not a spec
The 7,500 and 50 figures come from YouTube's documentation. Everything else is a curve fitted through them, and the million-point column is an arbitrary yardstick — YouTube publishes no leaderboard threshold, and the real cut-off moves every day with whatever else was uploaded that week in your country. The column is there to compare channel sizes against each other, not to give you a target.
What the curve actually does to a size advantage
The obvious reading is that the handicap levels the field. It does not. Work it through: if two channels convert the same fraction of their subscribers into hypes, total points scale as subscribers raised to the power of 0.27 — the size advantage survives, just heavily flattened.
Concretely, take a 2% conversion rate on both ends of the range. A 500-subscriber channel gets 10 hypes at 7,500 each: 75,000 points. A 500,000-subscriber channel gets 10,000 hypes at 50 each: 500,000 points. A channel a thousand times larger, doing exactly as well with its audience, still finishes about seven times ahead.
Seven times is a beatable gap, and that is the real design of the mechanism. It converts an impossible contest of scale into a winnable contest of intensity. To match that 500,000-subscriber channel at 2%, a 5,000-subscriber channel needs roughly 354 hypes — about 7% of its subscriber base. A hundredfold size disadvantage is erased by a three-and-a-half-fold advantage in how many of your viewers will actually press a button for you.
That reframes the whole question. Hype does not reward being small. It rewards having a disproportionately committed audience for your size, and it is one of the very few places on the platform where that quality is measured directly rather than inferred from watch time.
How much hyping actually happens
YouTube published one usable volume figure, from the beta. In the first four weeks across Brazil, Turkey and Taiwan, viewers hyped over 5 million times across more than 50,000 unique channels. The division is worth doing: that averages about 100 hypes per channel over four weeks, or roughly 25 a week.
Treat that number carefully. It is a mean, and hype counts almost certainly follow the same brutal distribution as everything else on YouTube, so the median channel got far fewer than 100 and a small number got thousands. It also comes from three countries during a novelty period, before the feature reached 39 markets. But as an order of magnitude it is clarifying: at the scale of an ordinary channel, hypes arrive in dozens, not thousands. Cross-reference that against the table above and the picture sharpens — dozens of hypes is a meaningful score at 1,000 subscribers and statistical noise at 200,000.
YouTube also reported that 18-to-24-year-olds made up over 30% of beta users, and that in a survey of 2,500 viewers aged 18 to 45 across the US, Japan and Germany, more than 75% — and more than 80% of Gen Z respondents — said they wanted to support small and medium creators. The appetite is real. The distribution of who acts on it is young and app-native, which matches a feature that exists only inside the mobile app.
The leaderboard is a browse surface with no personalisation
This is the part that matters most and gets the least attention. Points are not the prize; placement is. Tap the compass icon in the top-left of the mobile app, then Hype, and you reach the leaderboard: up to 50 hyped videos, filterable by topic, and — the crucial property — the same for every viewer in the same country. It is not personalised.
Every other browse surface on YouTube is a prediction about you specifically. The home feed is assembled per viewer; suggested videos depend on what you just watched; even search results are tilted by history. That selection is why a click-through rate from one surface means nothing next to the same figure from another — the audiences were filtered differently before they ever saw your tile. The Hype leaderboard applies no such filter. It is a shop window, identical for everyone who walks past.
Two consequences follow, and they point in opposite directions.
The first is that nobody browsing the leaderboard knows who you are. There is no "because you watched" behind your tile, no prior affinity, no channel recognition doing quiet work in the background. The audience arriving at your thumbnail is as cold as an audience gets on YouTube — closer to a search results page for a term you have never ranked for than to anything in the home feed.
The second is that, uniquely, you can see the competition in advance. Fifty tiles, refreshed continuously, filtered by the same topics you would be filed under. Open the leaderboard for your category before you upload and you are looking at a live sample of exactly what your artwork will sit beside if the hypes land. No other surface on the platform will show you that.
The badge sits on your thumbnail
A video that places on the leaderboard carries a Hyped badge and its leaderboard rank for seven days. That is a piece of interface furniture landing on top of artwork you designed without it — the same category of problem as the duration stamp in the bottom-right corner. If your composition puts anything load-bearing where a badge might sit, you lose it precisely when the most people are looking.
Designing for fifty tiles you can actually see
A leaderboard of fifty videos from one country, drawn from one week, is a much stranger grid than the home feed. The channels in it have nothing in common except size band and recent momentum. A cooking video, a Minecraft video and a video essay about aviation disasters can sit in consecutive rows.
That mixture changes what standing out means. In a niche feed, differentiation is a fight against near-identical conventions — every thumbnail in the row uses the same palette, so escaping it means breaking one. On the Hype leaderboard there is no shared convention to break. The competition is visual noise rather than visual sameness, and the thumbnails that survive noise are the ones with the simplest luminance structure: one subject, one message, hard separation between foreground and background. Our guide to choosing thumbnail colours covers the greyscale test that predicts this in about four seconds.
The cold-audience property compounds it. When nobody recognises the channel, everything the thumbnail usually gets to assume disappears. A face that means something to your subscribers means nothing here. An in-joke in the title lands on nobody. A running series number — Episode 47 — actively repels a viewer who has not seen the previous 46. The tiles that do well in front of strangers are the ones that state a complete proposition on their own, which is a genuinely different brief from the one you write for your subscriber feed.
It is worth checking your candidate against the grid rather than against your own canvas. Our thumbnail preview tool renders artwork at the sizes YouTube actually serves, which is the only condition under which the question "does this read?" has a real answer.
What Hype does not do
A great deal of what has been written about Hype describes it as an algorithmic boost. YouTube has never described it that way. What the documentation describes is a points total, a leaderboard, and a badge. There is no statement anywhere in the official material that hypes feed into recommendations, search ranking or the suggested-video system.
That does not mean placement is inert. Anything that generates real views generates real watch time, and watch time is a ranking input everywhere — so a video that lands on the leaderboard and holds viewers may well be recommended more afterwards. But that is an ordinary second-order effect of getting views, identical to what would happen if the views had come from a link on a forum. It is not evidence of a hidden hype signal, and any guide claiming a specific ranking uplift from hypes is describing something nobody has published. Our audit of the 2026 algorithm changes goes through the same exercise for a dozen other claims that circulate without a source.
Three more limits worth knowing. Hype pays the creator nothing directly — the free version is pure visibility. Creators cannot see who hyped a video; Studio reports the number of hypes and the points total, in a Hype card in analytics and in the weekly recaps, but not identities. And the badge and rank expire after seven days, along with the video's eligibility to be hyped at all.
How to ask without ruining it
Hype is the one lever that responds to asking, which makes it the one lever that can be spoiled by asking badly. Three constraints should shape how you do it.
It is scarce. Three hypes a week is a budget, and every creator your viewer follows who is under 500,000 subscribers is bidding for the same three. You are not asking for a free gesture; you are asking someone to prefer you over other channels they like. That argues for asking rarely and for asking at the moment when preference is highest — which is at the end of a video someone has actually watched, not in the first fifteen seconds before you have earned anything.
It is app-only. A viewer watching on a television or a laptop cannot act on your request at all, and telling half your audience to do something impossible is a small tax on your credibility. One clean sentence that names the constraint — that the button is in the app, under Like — does more than a paragraph of enthusiasm.
It is time-boxed. Seven days, and then the ability disappears. That makes Hype awkward for channels with a long tail and unusually well-suited to channels with a genuine release rhythm, where the audience has a habit of showing up early. If your uploads land irregularly, the window will routinely close before your viewers arrive; our analysis of upload frequency covers why cadence consistency is worth more than raw volume, and Hype is one more argument on that side of the ledger.
The mechanism does at least give you a place to point people. Since November 2025, viewers who hype a video can share it as a public post to their own channel's posts tab, which turns a private tap into something visible to their audience. That is the closest thing Hype has to a compounding effect, and it costs the viewer one extra step.
Paid hypes, and where this goes next
YouTube has been testing paid hypes in Brazil and Turkey — viewers buying additional hype points beyond their free weekly three. It is easy to be cynical about this, and the cynical reading is partly right: a leaderboard where points can be purchased is a leaderboard with a wallet-shaped distortion in it.
The more useful reading is that it makes Hype a fan-funding product rather than a discovery product. Alongside memberships, Super Thanks and Shopping — all of which open at the same 500-subscriber Partner Program tier — paid hype fits a pattern of YouTube building ways for small audiences to transfer value to creators directly, rather than waiting for advertising revenue that a small channel will never accumulate meaningfully. Whether that improves the leaderboard as a discovery surface is a separate question, and the honest answer is that nobody outside YouTube can tell yet.
Is it worth your attention?
Here is the honest verdict. If you are in the Partner Program and under about 20,000 subscribers, Hype is one of the highest-leverage things available to you, because the points curve is steep exactly where you are standing and the number of hypes required is a number your audience could plausibly produce. Between 20,000 and 100,000 it becomes a stretch that rewards an unusually engaged community. Above that, the arithmetic turns hostile: at 250,000 subscribers you need roughly 12,000 hypes to reach the same score a 1,000-subscriber channel reaches with 220, and you are competing against channels being handed a 90-fold multiplier.
What none of that changes is the second half of the job. Points decide whether your video appears on the leaderboard. They decide nothing at all about whether anyone taps it once it is there. Fifty tiles, no personalisation, no channel recognition, and a viewer who scrolls at the same speed they scroll everywhere else — that is a pure packaging test, run in public, against a competitive set you can actually go and look at.
Which is the recurring shape of every platform feature worth learning. Hype changes who gets shown. It does not change what gets clicked, and it never was going to. If the artwork is the constraint — and for most channels under 500,000 subscribers it is — then earning fifty-tile placement with a thumbnail that cannot hold its own in a grid of strangers just buys you a more public version of the same failure.
That is the part Thumblore exists for: producing artwork that reads at feed size and survives a row it did not choose, in the time it takes to write a description. Go and open the leaderboard for your category first, though. Fifty tiles, updated continuously, showing you exactly what you would be sitting next to — it is the most honest competitive research YouTube has ever made available, and almost nobody uses it.