Every perp venue that has already listed a token: what it launched at, how it has traded since, what points actually paid, and how the multiple has held up.
Median FDV at each point after listing, relative to day one. Launch valuations have historically compressed, which is why the model offers a decay haircut — the multiple a venue lists at is not the multiple it holds.
Where past tokens priced against the flow they were doing, with today's programmes plotted as forecasts on the same axes. Both scales are logarithmic and the trend line is fitted in log space.
| 2026-07-30 | 28.0% | 444.7M | $0.1461 | $231.99M | $235.12M | 101% | $360.00M | $1.20B | 0.00× | |
| 2026-03-31 | 30.0% | 7.31M | $4.61 | |||||||
| $780.89M | $214.89M | |||||
| $1.54B | $231.99M | +565% | $0.7432 | $0.1461 | +409% | |
| $4.67B | $673.38M | +594% | $146.66 | $4.61 | +3081% | |
Each launch is priced with multiples fitted on the other launches only, so the model never sees the answer it is being graded against. Median absolute error is +178% on FDV and +409% on value per point. Errors of this size are normal for pre-launch valuation and are precisely why the page shows a bear/bull band rather than a single number — treat any point estimate here as the middle of a wide distribution.
Each line is one token's price relative to its listing day, from daily closes on a venue that listed it from launch; where that venue came a few days late, the opening days are bridged from CoinGecko's FDV prints. The mean and band are geometric — the average and one standard deviation of log returns across the tokens listed that long — because the spread is multiplicative: a token at 3× and one at ⅓× are equally far from 100. Dashed: no exchange candles, so DIME is interpolated between the 0/7/30/90/180/365-day snapshots. Click a token to drop it from the chart and the statistics.