Solana's supply, inflation and staked share
How much SOL exists, how fast new SOL is created, and how much of it is staked, read from the network. One subtlety most pages miss: the inflation rate is per protocol year, and a protocol year is now longer than a calendar year.
637.62M SOL existed on November 16, 2026, 591.63M SOL of it circulating by the RPC's definition. The network reports inflation of 3.556% per protocol year, a year counted in slots; at the measured epoch length new SOL is being created at about 3.37% per calendar year. 67.3% of all SOL is staked. The schedule (8% at the start, cut 15% a year, 1.5% floor) is unchanged on-chain: the faster cut voted on in August 2026 is not active.
Supply
| Item | SOL | Share of total | Source |
|---|---|---|---|
| Total supply | 637,615,672 | 100.00% | getSupply |
| Circulating | 591,628,929 | 92.79% | getSupply |
| Non-circulating | 45,986,743 | 7.21% | getSupply |
| Staked (activated stake) | 429,328,621 | 67.33% | getVoteAccounts |
| Not staked | 208,287,051 | 32.67% | computed |
"Circulating" is the RPC's own split: it leaves out accounts the Solana Foundation lists as non-circulating. It is a bookkeeping category, not a measure of what trades. Total supply is net of burned fees and the Validator Admission Ticket, which destroy SOL.
Inflation: the network's figure and the calendar-year figure
- Per protocol year (network figure)
- 3.556%
- Per calendar year (computed)
- 3.37%
- Epochs a year (last 10 epochs)
- 340.1
- One protocol year lasts
- 392 days
Solana issues new SOL once per epoch. The protocol does not look at a clock to decide how much: it treats each epoch as a fixed fraction of a year counted in slots. When slots were cut to 200 ms, SIMD-0525 scaled that year to 157,784,629.968 slots "so inflation remains unchanged from a wall-clock perspective", which makes one 432,000-slot epoch exactly 1/365.24 of a year: a day, if slots really ran at 200 ms.
They do not. Slots run at about 215 ms, an epoch takes about 25.8 hours, and there are about 340 epochs in a calendar year, not 365. Each epoch still issues a day's worth. So per calendar year:
(1 + 3.556% ÷ 365.24)340.1 − 1 = 3.37% per calendar year (computed: each epoch adds one protocol day's issuance to the supply, compounded over the measured epochs in a year).
The same scaling applies to the staking reward: there are 6.9% fewer reward payouts in a calendar year than the protocol year assumes, so a staking yield worked out from the network's rate as if a year had 365 epochs overstates what a calendar year pays. The staking ledger avoids this by measuring each staking token's real gains per epoch and annualizing with the measured epoch length.
We checked it against the chain. A validator charging 100% commission receives all the inflation rewards its delegated stake earns, so its reward ÷ its stake is that epoch's reward rate. For the 5 largest such validators in epoch 1086, the reward was 1.444 × 10⁻⁴ of stake on average (1.43 × 10⁻⁴ to 1.46 × 10⁻⁴), against 1.446 × 10⁻⁴ predicted by the formula rate × 432,000 ÷ 157,784,629.968 × supply ÷ stake: a ratio of 0.999. Each epoch pays one protocol day of inflation, however long it actually took (getInflationReward, read November 16, 2026).
The schedule, and where it stands
getInflationGovernor returns the schedule's parameters: start at 8%, cut the rate by 15% of itself each year, stop at a floor of 1.5%. Today's 3.556% puts the network 4.99 protocol years into it. If nothing changes, the rate reaches the floor in about 5.3 protocol years.
| Protocol years from now | Network rate, per protocol year |
|---|---|
| now | 3.56% |
| 1 | 3.02% |
| 2 | 2.57% |
| 3 | 2.18% |
| 4 | 1.86% |
| 5 | 1.58% |
| 6 | 1.50% |
Two proposals to change it: SIMD-0228 (which would have tied the rate to how much SOL is staked) failed to reach quorum in 2025, as SIMD-0550's own text records. SIMD-0550, "double disinflation", would raise the yearly cut from 15% to 30%; crypto.news reported that a stake vote passed with 67% of participating stake on August 28, 2026. Its feature account does not exist on mainnet as of November 16, 2026, so the schedule above is the one in force. Our upgrade log checks for it every run. The history and the debate are explained in Solana inflation.
Staked share
429,328,621 SOL, 67.3% of all SOL, is staked (activated stake across every vote account). Because new SOL goes to stakers, the staked share is what decides how inflation divides: stakers' share of all SOL grows a little each year, and everyone else's shrinks by roughly the inflation rate. What that means in numbers.
Daily readings
| Date | Epoch | Total SOL | Staked | Inflation (network) | Per calendar year |
|---|---|---|---|---|---|
| 2026-11-16 | 1087 | 637,615,672 | 67.33% | 3.556% | 3.37% |
| 2026-11-15 | 1086 | 637,554,571 | 67.39% | 3.557% | 3.37% |
| 2026-11-14 | 1085 | 637,493,510 | 67.83% | 3.559% | 3.37% |
| 2026-11-13 | 1085 | 637,493,946 | 67.83% | 3.559% | 3.37% |
| 2026-11-12 | 1084 | 637,432,954 | 67.99% | 3.560% | 3.38% |
| 2026-11-11 | 1083 | 637,371,753 | 68.00% | 3.562% | 3.38% |
| 2026-11-10 | 1083 | 637,372,100 | 68.00% | 3.562% | 3.37% |
How this is measured
Staked share = activated stake of every vote account ÷ total supply. The network's inflation rate is per protocol year of 157,784,629.968 slots (SIMD-0525), so each 432,000-slot epoch issues 1/365.24 of the yearly rate. Rate per calendar year (computed, compounded per epoch, the site-wide definition in pipeline/datasets/_yield.mjs) = (1 + network rate × 432,000 ÷ 157,784,629.968)^N − 1, where N = 8,760 hours ÷ the mean length of the last 10 completed epochs (the same N on every page of this site). Years on the schedule = ln(rate ÷ initial) ÷ ln(1 − taper). Issuance check: for the five largest validators charging 100% commission, last epoch's inflation reward ÷ their activated stake, against the formula rate × 432,000 ÷ 157,784,629.968 × total supply ÷ total stake.
- "Circulating" is the RPC's own split: it excludes accounts the Solana Foundation lists as non-circulating. It is not a measure of what trades.
- The calendar-year rate depends on the measured epoch length; if slots speed up toward the 200 ms target, it moves toward the network's figure.
- The issuance check uses today's activated stake for last epoch's reward, so stake that moved in between adds a little noise.
- The schedule shown assumes no change. SIMD-0550 (a faster taper) passed a stake vote in August 2026 per press reports but has no feature account on mainnet, so it is not in effect.
- Burned fees and the Validator Admission Ticket remove some SOL; total supply is net of them.
Questions people ask
What is Solana's inflation rate?
About 3.37% per calendar year (computed from the measured epoch length, compounded per epoch). The network itself reports 3.556% per protocol year (getInflationRate, epoch 1087); that "year" is a fixed number of slots, and slots run slower than the 200 ms target.
What is the total supply of SOL?
637,615,672 SOL on November 16, 2026 (getSupply). The RPC counts 591,628,929 as circulating; the rest sits in accounts the Solana Foundation lists as non-circulating.
Will Solana inflation keep falling?
Under the current schedule the rate falls 15% each protocol year until it reaches the 1.5% floor, about 5.3 protocol years from now. SIMD-0550 would double the cut to 30% a year; a stake vote passed in August 2026 (reported), but it has no feature account on mainnet, so it is not in effect.
How much SOL is staked?
429,328,621 SOL, 67.3% of total supply, on November 16, 2026 (activated stake from getVoteAccounts).
Does unstaked SOL lose value from inflation?
Its share of all SOL shrinks: new SOL goes to stakers, so an unstaked holder's slice gets smaller by roughly the inflation rate each year. That is a change in share, not a price statement. What staking really pays after inflation.
Sources
SolPilled is independent and unofficial: not the Solana Foundation, Anza, or any validator. Figures are read from public endpoints at the time shown and can be wrong if an endpoint is; for the network's state right now, check status.solana.com and a block explorer. Nothing here is investment advice.