osu! rank tiers: what pp you need for each rank band
An osu! global rank is a position in a list ordered by total pp, not a fixed score threshold, so the pp required for any rank band shifts continuously as players gain pp, lose it to reworks, or expire from the leaderboard. Measured on 12 August 2026, about 13,799 pp placed a player at rank 1,000, about 9,476 pp at rank 10,000, about 5,100 pp near rank 100,000, and about 720 pp near rank 1,000,000. The curve steepens sharply toward the top: near rank 1,000 a hundred pp is worth roughly 75 places, while near rank 1,000,000 the same hundred pp is worth tens of thousands.
How the ranking is assembled
Your profile pp is a weighted sum of your ranked scores. Each score is multiplied by 0.95 raised to the power of its position minus one, so your best play counts in full, your tenth counts 63.02%, your twenty-fifth 29.20%, and your hundredth 0.6232%. The osu! website and API display the top 100 as your Best Performance, but the weighting itself does not stop there — scores beyond position 100 continue to contribute, simply at weights below 0.63%.
On top of that weighted sum sits bonus pp, awarded for the number of distinct ranked beatmaps you have set a score on. Global rank is then just your position once every ranked player is sorted by that combined total.
Nothing about a rank is a fixed requirement. It is entirely relative to everyone else's totals at that moment, which is why every number in this guide carries a measurement date.
pp by rank band, measured on 12 August 2026
Ranks 1 to 10,000 below are read directly from the official osu! performance rankings, which is the deepest the public rankings expose. Deeper figures are interpolated between live player profiles sampled on the same day and rounded, so they are approximate by construction. Treat the whole table as a dated snapshot rather than a specification.
| Global rank | Approximate total pp |
|---|---|
| 1 | 32,139 |
| 10 | 24,057 |
| 50 | 20,042 |
| 100 | 18,677 |
| 250 | 16,624 |
| 500 | 15,132 |
| 1,000 | 13,799 |
| 5,000 | 10,701 |
| 10,000 | 9,476 |
| 25,000 | ~7,700 |
| 50,000 | ~6,400 |
| 100,000 | ~5,100 |
| 200,000 | ~3,700 |
| 300,000 | ~2,900 |
| 500,000 | ~1,900 |
| 750,000 | ~1,200 |
| 1,000,000 | ~720 |
| 1,500,000 | ~240 |
Two cautions on reading it. Registered osu! accounts vastly outnumber ranked ones, because holding a rank requires recent activity, so the ranked population is far smaller than the total playerbase. And the spacing between rows is deliberately uneven: the top of the table covers a few hundred players across thousands of pp, while the bottom rows each span hundreds of thousands of players across a few hundred pp. Reading the rank column as though it were linear will badly mislead you about how far apart two players actually are.
Why the curve steepens near the top
The pp axis is stretched at the top and compressed at the bottom because players are not spread evenly along it. Most of the population sits within a narrow pp range, so a small pp change crosses many ranks. At the top, few players occupy a wide pp range, so large pp changes barely move you.
| Around rank | pp per 1,000 ranks | Ranks gained per 100 pp |
|---|---|---|
| 1,500 | ~1,341 | ~75 |
| 7,000 | ~245 | ~410 |
| 115,000 | ~17.6 | ~5,700 |
| 500,000 | ~3.8 | ~26,000 |
| 1,000,000 | ~1.5 | ~67,000 |
The practical reading: at six-digit ranks, modest pp gains produce dramatic rank jumps, which is why early progress feels rapid. Near the top, a player can add several hundred pp across a season and move a handful of places. Neither fact says anything about rate of improvement — they are two ends of the same distribution.
What each band means in playing terms
A useful way to translate a rank into something actionable is to ask what your top plays would have to average. If a top 100 were perfectly flat, with every play worth the same raw pp, the hundred weights sum to 19.88, so the total is roughly 19.88 times that average plus bonus pp. Working backwards, assuming a full bonus:
| Global rank | Approx. total pp | Implied flat top-100 average |
|---|---|---|
| 1,000 | 13,799 | ~673 pp per play |
| 10,000 | 9,476 | ~456 pp per play |
| 100,000 | ~5,100 | ~237 pp per play |
| 500,000 | ~1,900 | ~75 pp per play |
| 1,000,000 | ~720 | ~16 pp per play |
Real top-100 lists are not flat — they slope downwards, so an actual best play sits well above these averages and the hundredth well below. The model is still the right shape for target-setting, because it tells you whether you need one better score or a deeper set of them. Community shorthand names these bands by digit count: four-digit means ranks 1,000 to 9,999, five-digit means 10,000 to 99,999, and so on.
Bonus pp: the part of the total you cannot farm indefinitely
Bonus pp is awarded for the number of distinct ranked beatmaps you have scored on, following 416.6667 × (1 − 0.995min(N, 1000)), where N is that count. It reaches half its value at roughly 137 scores and tops out at 413.894 pp at 1,000 scores, after which it cannot grow. Note this counts distinct ranked maps, not your play count — replaying the same beatmap adds nothing to it.
Because the ceiling is fixed, its significance collapses as you climb:
| Global rank | Total pp | Full bonus as share of total |
|---|---|---|
| 1 | 32,139 | 1.3% |
| 1,000 | 13,799 | 3.0% |
| 10,000 | 9,476 | 4.4% |
| 100,000 | ~5,100 | 8.1% |
| 500,000 | ~1,900 | 21.7% |
| 1,000,000 | ~720 | 57.3% |
For a newer player, spreading scores across many ranked maps is among the fastest pp available — and it is finite. These shares assume a full bonus; a player with fewer than 1,000 ranked scores has less.
Why the same pp buys a worse rank over time
Three separate forces move the pp-to-rank mapping, and only one of them is you.
- Population growth and collective improvement. As more players reach a given pp total, that total slides down the list. This is the dominant effect across five- and six-digit ranks.
- pp reworks. The algorithm is revised periodically and scores are recalculated. The January 2021 revision changed approach-rate scaling and miss handling among other things; the October 2024 revision removed combo scaling in osu!, reworked rhythm complexity, adjusted fast flow aim, and let sliders contribute to accuracy pp for scores using slider head accuracy. Totals shift for the whole playerbase at once when this happens.
- Leaderboard eligibility. Players who stop submitting scores are eventually treated as inactive and drop off the ranked leaderboards, lifting everyone below them. They reappear once they submit again.
This is why any pp-to-rank table needs a date attached to be meaningful. The figures here describe 12 August 2026 and will drift from that day onwards.
Turning a rank goal into a play goal
Work in three steps. Convert the target rank into a pp total using a current table, subtract your present pp to get the gap, then convert that gap into scores using the weighting curve.
Worked example. A player at 5,100 pp sits near rank 100,000 and wants rank 50,000, which needed roughly 6,400 pp on 12 August 2026 — a gap of about 1,300 pp. That gap cannot be closed by one score. A new play entering at position 1 contributes its full value, but one entering at position 20 is multiplied by 0.377, so a 400 pp play in that slot adds only about 151 pp before anything else is accounted for.
The weighting also charges you for displacement: every play pushed down a slot loses 5% of its weighted contribution, so the net gain is always smaller than the headline pp of the new score. Practically, closing a 1,300 pp gap at that level means raising the whole upper half of the list rather than landing one exceptional play.
Tracking your own movement against the drift
Because rank is relative, it changes on days you do not play. Other players gaining pp push you down; inactive accounts dropping off lift you up. A rank that slipped overnight is not evidence that you got worse. Country rank moves against a much smaller population and is correspondingly noisier, sometimes jumping several places from a single other player's session.
The way to separate signal from drift is to record pp and rank together over time rather than checking either in isolation. TrackMyOsu snapshots pp, global rank and country rank when you start a session and again when you stop, keeps every session permanently, and flags any new top plays and medals detected while it runs.
Across a run of sessions that distinguishes genuine pp gains from leaderboard noise, and the pp-gains leaderboard and Hall of Fame put those gains in context against other tracked players.
By carrot12_. Last updated 2026-08-12.