Everything we compute from the Base blockchain, on one page. Every point comes from a direct call to the contract or from its events, not from an aggregator. Hover a line to see the value on a given date; the period switches with the buttons.
The same series are served to machines as JSON — see the API.
Pool size by day since the contract's first day, 31 March 2026. Every point is a totalPooledRnbw() call at a historical block; two independent nodes returned the same result.
Growth was not continuous. The sharp rise in April was the start of the allocation programme: in two weeks the pool went from 200 thousand to 5.7 million. The falling days are the ones where exits outweighed entries.
Series: /api/tvl.json
The number of addresses with an open position, by day. It is computed from contract events: an entry adds an address, an exit removes it. Partial exit is disabled in the contract, so Unstaked always means a full exit and the count is exact rather than an estimate.
Do not confuse this with pool size. The number of addresses can rise while the number of tokens in the pool falls — if large holders leave and small ones arrive. Put the two charts side by side and you can see who the pool is filling up with.
Series: /api/stakers.json, column 1. Built from every Staked and Unstaked event of the contract.
The number of wallets with a non-zero token balance. The series is reconstructed from every transfer of the token since its first day rather than copied from an explorer: an address balance is everything received minus everything sent. The current figure is taken hourly from GeckoTerminal.
The 250 addresses with the largest balance, with labels taken from the explorer — it knows exchanges, pools and Rainbow's own contracts by name. Together they hold almost the entire supply, so the other twenty-five thousand addresses no longer change the picture. The full list cannot be pulled: the explorer serves fifty addresses per request and shuts down for hours after about fifty requests.
The history is complete — from the token’s first day. What used to be here was a series starting the day we began recording an explorer’s reading, plus a caveat that the past had not been recomputed yet. Now it has: all 1,225,924 transfers since 26 January were read, and the sum of the reconstructed balances together with what was burned matched the supply down to the wei. Holders peaked on 13 April at 28,027 addresses; there are 9% fewer now.
A falling line does not mean people are leaving. Staking takes the tokens into the contract, so an address that locks its whole balance goes to zero and drops out of the holder count. Over the last 30 days the pool grew by 7.31 million RNBW while the holder count fell by 1,022 — part of that fall is a move into staking, not a sale. The count also includes every address alike, contracts included: the liquidity pool, the staking contract itself and the treasury sit in it next to people.
Series: /api/holders.json. Reconstructed from the token’s transfers; the current figure comes from GeckoTerminal. Blockscout’s counter used to be here; it overstated the holder count by roughly a factor of two (53,678 against an actual 25,402 on 5 October) and has since closed behind a Cloudflare challenge.
The annual rate computed from the exit fees and allocations actually recorded over the last week, not from promises. It jumps not because something is broken: rewards are split among everyone in the pool, so the rate rises on exits and falls on entries.
Series: /api/apr.json, column 1. Reconstructed from the history of exit fees and daily pool size.
The same thing on a monthly window. The weekly line shows mood, the monthly one shows the trend: on it you can see where the rate is heading once the noise averages out.
Series: /api/apr.json, column 2.
Every exit from staking leaves 10% of the position in the contract. That is where the drip comes from: each bar is the sum of all exit fees for a day. Bars rather than a line, because this is not a level that holds but events on particular days.
Where those coins go. Not to the project and not to a burn: what is collected drips evenly over seven days to everyone still in the pool. So a tall bar here is somebody's mass exit and, at the same time, a week of higher yield for everyone else.
Series: /api/pot.json, column 1. Exit fees have been recorded since 16 May 2026 — whatever was collected before that date is not in the series.
The same figure as a running total. The slope of the line shows how actively people are leaving: a flat stretch is a calm period, a steep one is a wave of exits.
Series: /api/pot.json, column 2.
How many dollars a day pass through the main pool on Uniswap v4. This is what liquidity providers' income is made of — and also a measure of how much the token trades at all.
Series: /api/pool.json, column 1. GeckoTerminal data, polled every ten minutes.
The pool's rate is 1% of volume, and that is what liquidity providers divide among themselves. Compare it with the pool size on the Earn page to get the yield: that is exactly how it is calculated, rather than taken from an announcement.
Do not forget impermanent loss. Fees are income but not the whole result: when the price moves, the composition of your share in the pool changes, and the outcome can be worse than simply holding the tokens. Explained on the Earn page.
Series: /api/pool.json, column 2.
One permanent connection to a Base node for all visitors. Otherwise every person opening the page would open their own subscription, and we would hit the public nodes' limits at the first dozen readers.
Fifty lines of our own code instead of someone else's script from an external domain. A side benefit: the colours come from the same variables as the rest of the page, so light and dark themes work by themselves.
We do not draw in history that does not exist. The holder count starts today, exit fees start on 16 May, and pool size starts on the contract's first day. Every chart is labelled with its own source.
The series sit there as ready JSON, with no keys and open CORS. The list and the format are on the API page.