BNB Tokenomics Explained
BNB (Binance Token) has one of the longest-running deflationary supply programs in crypto. It launched in 2017 with 200 million tokens and is systematically burning down to a permanent floor of 100 million. As of the July 2026 quarterly burn, roughly 133 million remain.
That makes it a useful case study, because most tokens describe a burn mechanism in a whitepaper and never execute it consistently. BNB has run the program for years, in public, on a schedule.
The structure
BNB's supply reduction runs through two separate mechanisms working at different speeds.
Auto-Burn is the quarterly event. Introduced in late 2021, it replaced the earlier system where Binance burned tokens based on its own exchange profits. The current version uses a published formula based on the average BNB price over the quarter and the number of blocks produced on the chain, which produces a fixed burn amount at regular intervals.
The formula has a counter-cyclical property worth noticing: when the price is lower, more tokens get burned. When the price is higher, fewer units are burned even though the notional dollar value stays large. The mechanism targets a supply schedule rather than a spending amount.
BEP-95 is the continuous one. It burns a portion of gas fees in real time, block by block, with an initial burn ratio set at 10%. Unlike Auto-Burn, this ties supply reduction directly to network usage: more activity on the chain means more BNB destroyed, permanently, with no quarterly event required.
The distinction matters for anyone designing their own tokenomics. One mechanism is scheduled and predictable. The other is usage-linked and variable. Running both means the supply curve isn't dependent on a single input.
Where the numbers actually stand
The 36th quarterly burn executed on 15 July 2026 destroyed 1,615,827.795 BNB, worth approximately $931.7 million at the time, bringing total supply to roughly 133.17 million.
Cumulative burns since launch total approximately 67.4 million BNB. That leaves roughly 33.2 million to reach the 100 million floor. At current rates, that target lands somewhere around 2027 to 2028.
What this design does well
It's verifiable. Every burn is on-chain and the formula is published. Anyone can check the claim against the ledger, which is a materially different thing from a project stating a deflationary intention in a document.
It has a defined endpoint. The 100 million floor is a commitment to stop, not an open-ended burn. That gives holders a supply figure to model against rather than an indefinite reduction.
It doesn't depend on one input. Scheduled burns and usage burns fail in different conditions. Running both means a quiet quarter on-chain doesn't halt supply reduction entirely.
What it doesn't do
Worth stating plainly, because burn mechanics get oversold across the industry: reducing supply is not the same as creating demand.
A burn removes tokens from circulation. Whether that matters depends entirely on whether anyone wants the remaining ones. BNB's burn program operates alongside an exchange and a chain with real usage, and that usage is doing at least as much work as the burn. A token with an identical burn schedule and no demand would simply have fewer tokens that nobody wants.
This is the most common mistake in tokenomics design: treating a burn as a strategy rather than a supply-side tool. The tokenomics fundamentals that decide whether a token survives are allocation, vesting, and genuine demand drivers. Burns adjust supply. They don't substitute for the rest.
What founders can take from it
Three transferable lessons, none of which require running a chain:
Publish the formula. BNB's burns are credible because the calculation is public and checkable. A discretionary burn announced when convenient carries far less weight than a mechanical one.
Define the endpoint. An open-ended burn is a promise. A stated floor is a commitment with a number attached.
Link supply reduction to something real. BEP-95 ties burning to actual network usage. Any mechanism tied to genuine activity is harder to game and easier to defend than one tied to a team decision.
For the full framework on designing token supply, allocation and vesting, see our complete tokenomics guide.
FAQ
What is BNB's total supply?
BNB launched in 2017 with 200 million tokens. Following the July 2026 quarterly burn, total supply stands at approximately 133.17 million. The program targets a permanent floor of 100 million.
How does the BNB burn work?
Two mechanisms run in parallel. Auto-Burn executes quarterly using a published formula based on the average BNB price over the quarter and the number of blocks produced. BEP-95 burns a portion of gas fees in real time, block by block, at an initial ratio of 10%.
How many BNB have been burned so far?
Approximately 67.4 million BNB cumulatively since launch. The 36th quarterly burn, on 15 July 2026, destroyed 1,615,827.795 BNB, worth roughly $931.7 million at the time.
When will BNB reach 100 million supply?
At current burn rates, roughly 33.2 million BNB remain to be burned, which puts the target somewhere around 2027 to 2028. The exact timing depends on price and network activity, since both feed the burn calculations.
Is BNB deflationary?
Yes, in the sense that supply is reducing over time through both scheduled and usage-based burns. It's worth noting that the program has a defined stopping point at 100 million, so it's a reduction to a floor rather than indefinite deflation.
Does burning tokens increase the price?
Burning reduces supply, which is one side of the equation. It does not create demand for the remaining tokens. Whether a burn affects price depends on whether demand holds or grows as supply falls, which is a separate question from the burn mechanism itself.