The short answer: Ethereum has roughly 120 million ETH in existence, but there is no permanent “final number” the way there is with Bitcoin’s 21 million cap.
The exact ETH supply changes block by block.
New ETH is issued to validators for securing the network. Some ETH is permanently destroyed through fee burns. Staked ETH is still part of supply. Lost ETH is still counted. Wrapped ETH, liquid staking tokens, and bridged versions do not create new native ETH.
If you need the precise number right now, use a live supply tracker such as Etherscan or ultrasound.money and record the timestamp or block number. A static article can give the framework; the chain gives the current count.
How many Ethereum coins are there right now?
There are about 120 million ETH in total supply.
That number is intentionally approximate because Ethereum’s supply is not fixed at a single long-term maximum. It moves as three forces interact:
| Supply force | Effect on ETH supply | What causes it |
|---|---|---|
| Validator issuance | Increases supply | ETH paid to proof-of-stake validators |
| Fee burns | Decreases supply | Base fees burned under EIP-1559 and blob fee burns under EIP-4844 |
| Penalties and slashing | Slightly decreases supply | Validator penalties for being offline or misbehaving |
The practical answer depends on what you mean by “how many Ethereum coins”:
| Metric | What it means | Best use case |
|---|---|---|
| Total ETH supply | All native ETH that exists on Ethereum, including staked ETH | Monetary analysis, supply tracking |
| Circulating ETH supply | ETH considered available in market supply, depending on data provider methodology | Market cap comparisons |
| Staked ETH | ETH deposited to validators or staking services | Network security and staking analysis |
| Liquid staking tokens | Tokens like stETH or rETH representing claims on staked ETH | DeFi accounting, not native ETH supply |
| Wrapped ETH | ERC-20 representation of ETH, usually WETH | DeFi usage, not additional ETH |
| Bridged ETH | ETH representation on another chain or layer 2 | Cross-chain accounting, not new native ETH |
For most readers, the clean answer is:
Ethereum has roughly 120 million native ETH, and the exact amount changes continuously because ETH issuance and ETH burns happen as the network runs.
Why doesn’t Ethereum have a fixed coin supply like Bitcoin?
Ethereum and Bitcoin optimize for different monetary designs.
Bitcoin has a hard-coded maximum of 21 million BTC. Ethereum does not have a fixed maximum supply. Instead, ETH uses a dynamic supply model: issuance rewards validators, while network activity can burn ETH.
That does not mean ETH supply is unlimited in the practical sense. It means Ethereum does not promise a fixed terminal cap.
| Feature | Bitcoin | Ethereum |
|---|---|---|
| Maximum supply | 21 million BTC | No fixed maximum |
| New issuance | Mining subsidy, roughly every 10 minutes | Validator rewards, every epoch |
| Issuance trend | Programmed halvings | Depends on amount staked |
| Fee burn | No protocol-level fee burn | Base fee burns under EIP-1559 |
| Security model | Proof of work mining | Proof of stake validation |
| Monetary narrative | Hard cap scarcity | Dynamic issuance minus burn |
| Supply predictability | Very high | High in mechanism, variable in outcome |
The key difference is philosophical and technical.
Bitcoin minimizes monetary discretion by setting a fixed endpoint. Ethereum tries to balance three needs:
- Pay validators enough to secure the network
- Reduce unnecessary issuance after proof of stake
- Burn part of transaction fees so network usage can offset issuance
This is why Ethereum can be inflationary during quiet periods and deflationary during busy periods.
What actually changes the ETH supply?
ETH supply changes only when the protocol creates ETH, destroys ETH, or penalizes validator balances.
Many things that look like “supply changes” are not supply changes at all.
What increases ETH supply?
ETH supply increases through proof-of-stake issuance.
Validators earn ETH for performing duties such as:
- Attesting to blocks
- Proposing blocks
- Participating in sync committees
- Helping the chain reach consensus
This issuance is new ETH created by the protocol.
A subtle point: as more ETH is staked, total validator issuance generally rises, but the yield per validator falls. Ethereum is not paying a fixed interest rate. The reward curve adjusts based on total active stake.
What decreases ETH supply?
ETH supply decreases mainly through fee burns.
Since EIP-1559, every Ethereum transaction includes a base fee. That base fee is burned, meaning it is permanently removed from supply.
After EIP-4844, Ethereum also introduced blob transactions for rollups. Blob base fees are also burned.
ETH can also be reduced through validator penalties and slashing, although those are usually much smaller than fee burns during normal network conditions.
What does not change ETH supply?
This is where many supply discussions go wrong.
| Event | Does it change ETH supply? | Why |
|---|---|---|
| Sending ETH to another wallet | No | ETH moves between addresses |
| Buying ETH on an exchange | No | Ownership changes, supply does not |
| Staking ETH | No | ETH is locked or delegated, not destroyed |
| Unstaking ETH | No | ETH becomes liquid again, not newly created |
| Wrapping ETH into WETH | No | WETH is a representation of ETH |
| Bridging ETH to an L2 | No | ETH is locked or represented elsewhere |
| Minting stETH, rETH, or cbETH | No | These are claims on staked ETH |
| Losing a private key | No | The ETH is inaccessible, but still counted |
| Burning base fees | Yes | ETH is permanently removed |
| Validator rewards | Yes | New ETH is issued |
The clean mental model:
Staking changes liquidity. Wrapping changes token format. Bridging changes location. Only issuance and burns change native ETH supply.
How does ETH burning work?
ETH burning is easiest to understand through transaction fees.
Before EIP-1559, users paid gas fees to miners. After EIP-1559, every transaction has two main fee components:
| Fee component | Who receives it? | Does it affect supply? |
|---|---|---|
| Base fee | Burned by the protocol | Reduces ETH supply |
| Priority fee / tip | Validator or block builder path | Transfers ETH, does not reduce supply |
A user may feel like “the whole gas fee is burned,” but that is not correct. Only the base fee is destroyed.
Example: a normal ETH transfer
Suppose you send ETH on mainnet and the transaction uses:
- 21,000 gas
- 30 gwei base fee
- 2 gwei priority fee
The burn is:
21,000 × 30 gwei = 630,000 gwei = 0.00063 ETH burned
The priority fee is:
21,000 × 2 gwei = 42,000 gwei = 0.000042 ETH paid as a tip
Only 0.00063 ETH reduces supply.
The tip is not new ETH and not burned. It is a transfer from the sender to the validator/block production supply chain.
Example: a $100 USDT swap on Ethereum
Say a user swaps $100 USDT for ETH through a DEX on Ethereum mainnet.
The swap might use 120,000–180,000 gas depending on routing, token approvals, pool type, and contract path.
If the base fee is high, the burn can be meaningful relative to the trade size. If the base fee is low, the burn is smaller.
But the DEX trading fee is not the same as the ETH burn.
| Cost type | Example | Supply impact |
|---|---|---|
| Ethereum base fee | Burned gas fee | Reduces ETH supply |
| Priority fee | Tip for inclusion | No net supply change |
| DEX liquidity fee | Paid to liquidity providers | No net supply change |
| Price impact | Worse execution price | No net supply change |
A small trader often cares more about gas cost than supply economics. The protocol, however, only burns the base fee portion.
Example: a $10,000 swap with smart routing
A larger trader swapping $10,000 may route across multiple liquidity pools to reduce price impact. That can improve execution but may use more gas.
Platforms such as switchfi.app automatically compare multiple liquidity sources before selecting an execution route. From a supply perspective, the important detail is not which route is chosen; it is how much Ethereum gas is consumed and what the base fee is at execution time.
A route with better price execution can still burn more ETH if it requires more contract calls.
That trade-off matters:
| Route type | Likely gas use | Likely price impact | ETH burn effect |
|---|---|---|---|
| Simple single-pool swap | Lower | Potentially higher | Lower burn |
| Multi-hop route | Medium | Often lower | Medium burn |
| Split route across pools | Higher | Often best for large trades | Higher burn |
| L2 swap | Usually much lower user fee | Depends on L2 liquidity | Burn occurs partly through L1/blob settlement, not the same as mainnet |
For the trader, execution quality may matter more than burn. For ETH supply, network fee mechanics matter more than trade size.
Does staking reduce Ethereum supply?
No. Staking does not reduce ETH supply.
Staked ETH still exists. It is part of total supply whether it is held by:
- A solo validator
- A staking pool
- A centralized exchange staking service
- A liquid staking protocol
- An institutional validator setup
The confusion comes from liquidity.
If 30 million ETH is staked, that ETH may be less immediately available for trading, but it has not been burned. It can still earn rewards, be withdrawn under protocol rules, or be represented by liquid staking tokens.
Staked ETH versus liquid staking tokens
Liquid staking makes this even more confusing.
If someone deposits ETH into a liquid staking protocol and receives a staking token, that does not double the supply of ETH.
| Asset | What it represents | Is it native ETH supply? |
|---|---|---|
| ETH | Native Ethereum asset | Yes |
| stETH | Claim on staked ETH plus rewards model | No |
| rETH | Claim on Rocket Pool staked ETH | No |
| cbETH | Coinbase wrapped staked ETH representation | No |
| wstETH | Wrapped version of stETH | No |
Liquid staking tokens can affect DeFi liquidity, leverage, collateral markets, and risk. They do not create more native ETH.
Why staking can still affect market supply
Even though staking does not reduce total supply, it can affect liquid supply.
If a large share of ETH is staked or used as collateral, less ETH may be readily available on exchanges. That can influence market behavior, but it is not the same as protocol-level scarcity.
A useful distinction:
- Total supply answers: “How much ETH exists?”
- Liquid supply answers: “How much ETH is readily available to trade?”
- Free float answers: “How much ETH is likely to move in markets?”
- Effective supply answers: “How much ETH is available after accounting for staking, long-term holders, lost keys, and DeFi collateral?”
Most supply trackers focus on total supply because it is measurable. Effective supply is more interpretive.
Why did Ethereum become deflationary after the Merge?
Ethereum’s supply dynamics changed dramatically after the Merge in September 2022.
Before the Merge, Ethereum paid proof-of-work miners. After the Merge, Ethereum switched to proof of stake and stopped issuing miner block rewards.
That reduced new ETH issuance substantially.
At the same time, EIP-1559 continued burning base fees. During periods of high network activity, burns can exceed validator issuance. When that happens, ETH supply decreases.
Before and after the Merge
| Period | Main issuance source | Burn mechanism | Supply tendency |
|---|---|---|---|
| Pre-EIP-1559 | Miner rewards | None | Inflationary |
| EIP-1559 before Merge | Miner rewards | Base fee burn | Usually inflationary, but lower net issuance |
| Post-Merge | Validator rewards | Base fee burn | Can be inflationary or deflationary |
| Post-Dencun | Validator rewards | Base fee + blob fee burn | More dependent on L1 demand and blob fee markets |
The Merge did not create the burn. EIP-1559 created the main burn mechanism. The Merge reduced issuance enough for burns to sometimes dominate.
That distinction matters because many people say “the Merge made ETH deflationary.” More precisely:
The Merge sharply reduced issuance. EIP-1559 burns fees. Together, they made deflation possible.
Can Ethereum supply increase again?
Yes.
ETH can be deflationary, but it is not guaranteed to remain deflationary every day, month, or year.
Supply increases when validator issuance exceeds burns. That can happen when:
- Mainnet transaction demand is low
- Base fees remain cheap for long periods
- More activity moves to layer 2s
- Blob fees remain low
- Total validator issuance exceeds fee burn
- MEV and priority fees transfer value but do not burn ETH
This is not necessarily a failure. Lower fees are good for users. But lower base fees usually mean less ETH burned.
Ethereum’s scaling roadmap creates a trade-off:
| Goal | User benefit | Supply effect |
|---|---|---|
| Lower transaction fees | Better accessibility | Less base fee burn per transaction |
| More L2 usage | Cheaper execution | L1 burn depends on settlement/blob demand |
| Higher L1 congestion | More ETH burned | Worse user experience |
| More staking participation | Stronger validator set | Higher total issuance, lower individual staking yield |
| More efficient blockspace | Better throughput | Burn depends on market demand for blockspace |
A healthy Ethereum ecosystem does not require constantly high gas fees. But ETH supply bulls often prefer high burn. Those incentives are not always aligned.
How do layer 2 networks affect ETH supply?
Layer 2 networks change where users transact, but they still rely on Ethereum for settlement and data availability in different ways.
After EIP-4844, rollups can use blob space to post data more cheaply. This improved L2 economics but changed the fee burn profile.
Mainnet transaction versus L2 transaction
| User action | Where execution happens | What ETH burn depends on |
|---|---|---|
| ETH transfer on Ethereum mainnet | L1 | Mainnet base fee × gas used |
| DEX swap on Ethereum mainnet | L1 | Mainnet base fee × contract gas used |
| Swap on Arbitrum, Optimism, Base, or other L2 | L2 | L2 fees plus L1/blob settlement costs |
| Cross-chain transfer through a bridge | Multiple systems | Source/destination chain fees and settlement costs |
| Rollup batch posted to Ethereum | L1 data/blob market | Blob base fee or calldata base fee |
A user swapping on an L2 may pay much less than on mainnet. That is good for adoption. But the ETH burned per individual transaction is usually not equivalent to doing the same transaction directly on Ethereum mainnet.
This creates a more nuanced supply model:
- If L2 usage explodes and blob space becomes scarce, blob fees can burn more ETH.
- If L2 usage grows but blob capacity remains plentiful, user activity may not translate into large ETH burns.
- If mainnet becomes a high-value settlement layer, ETH burn may depend more on institutional settlement, rollup demand, and MEV-heavy activity than simple retail transfers.
The old mental model was “more Ethereum users equals more gas burned.” The newer model is closer to:
More demand for Ethereum blockspace and blob space equals more ETH burned.
Those are related, but not identical.
How should you verify the current ETH supply?
For casual reading, “about 120 million ETH” is enough.
For reporting, accounting, research, or trading analysis, use a live source and record the date, time, and block number.
| Source | What it is useful for | Strength | Limitation |
|---|---|---|---|
| Etherscan | Quick total supply checks | Widely used Ethereum block explorer | Depends on explorer methodology and indexing |
| ultrasound.money | Issuance, burn, and supply trend visualization | Excellent for monetary dashboarding | Dashboard values can differ slightly from other trackers |
| CoinGecko | Market cap and circulating supply context | Easy for market comparisons | Market data methodology may not match protocol accounting |
| Running/indexing your own node data | Research-grade verification | Maximum control | Technically demanding; total supply is not a simple one-call metric |
For serious work, do not just copy a number. Capture the context.
A good supply citation includes:
- Total ETH supply
- Source
- Timestamp
- Ethereum block number or epoch
- Whether the number is total supply or circulating supply
- Any methodology note if available
Example format:
ETH total supply: approximately 120.x million ETH
Source: Etherscan / ultrasound.money
Timestamp: YYYY-MM-DD HH:MM UTC
Block: [block number]
Metric: total native ETH supply
That level of precision prevents confusion later, especially because ETH supply can drift over time.
Why do different websites show different ETH supply numbers?
Different data providers may show slightly different ETH supply figures because they use different definitions and indexing methods.
Common reasons include:
| Reason | Why it matters |
|---|---|
| Total supply vs circulating supply | These are not always the same metric |
| Update frequency | Some dashboards update faster than others |
| Execution and consensus layer reconciliation | Post-Merge supply accounting spans both layers |
| Treatment of burned ETH | Burn addresses and protocol burns must be handled correctly |
| Staking rewards and withdrawals | Validator balances affect supply calculations |
| Rounding | Some sites round to the nearest thousand or million |
| Market data methodology | Coin data sites may prioritize market cap consistency |
Small discrepancies are normal. Large discrepancies are a red flag.
If one source says roughly 120 million ETH and another says 300 million, the second source is almost certainly wrong or measuring something else.
What is the difference between ETH, Ethereum, and “Ethereum coins”?
People often search for “Ethereum coins,” but the correct asset name is ether, ticker ETH.
Ethereum is the network. ETH is the native asset used to:
- Pay gas fees
- Stake validators
- Secure the network
- Settle transactions
- Serve as collateral in DeFi
- Act as the base asset across much of the Ethereum economy
The phrase “Ethereum coin” usually means ETH, but it can be ambiguous because Ethereum also hosts thousands of tokens.
| Term | Meaning |
|---|---|
| Ethereum | The blockchain network and protocol ecosystem |
| Ether | The native asset of Ethereum |
| ETH | The ticker symbol for ether |
| ERC-20 token | A token issued on Ethereum, such as USDC, UNI, or LINK |
| WETH | Wrapped ETH used as an ERC-20-compatible version of ETH |
If someone asks how many Ethereum coins exist, the answer should refer to ETH supply, not the number of ERC-20 tokens on Ethereum.
Does lost ETH reduce the supply?
Lost ETH does not reduce the official supply.
If someone loses a private key, sends ETH to an inaccessible address, or dies without passing on wallet access, that ETH may be economically gone. But the protocol cannot know whether the owner is truly gone or simply inactive.
So lost ETH remains counted in total supply.
This creates a gap between on-chain supply and economically available supply.
| Category | Counted in total supply? | Likely available to market? |
|---|---|---|
| ETH in active wallets | Yes | Maybe |
| ETH on exchanges | Yes | Often |
| Staked ETH | Yes | Partly, depending on withdrawal and holder behavior |
| ETH in DeFi contracts | Yes | Depends on collateral and lockup |
| Lost ETH | Yes | Probably not |
| ETH sent to burn addresses intentionally | Usually yes unless protocol-burned | Usually not |
| Protocol-burned ETH | No | No |
Burning through EIP-1559 is different from sending ETH to a famous dead address. Protocol burns reduce accounted supply. Lost keys usually do not.
What are the pros and cons of Ethereum’s dynamic supply model?
Ethereum’s monetary policy is harder to explain than Bitcoin’s, but it gives the protocol flexibility.
Pros
- Security funding adapts to staking participation. Validators are paid for securing the network.
- Network usage can offset issuance. High demand can reduce or even shrink supply.
- No proof-of-work miner subsidy is required. Post-Merge issuance is much lower than before.
- Fee burn aligns ETH with blockspace demand. Users competing for blockspace remove ETH from supply.
- Lower fees can coexist with monetary policy. The system does not require a permanent high-fee environment, though burns fall when fees are low.
Cons
- Harder to communicate. “About 120 million and changing” is less simple than “21 million.”
- Supply forecasts depend on usage. Future ETH supply depends on fees, staking, L2 demand, and protocol changes.
- Deflation is not guaranteed. Quiet network periods can make ETH inflationary again.
- Scaling can reduce burn pressure. Cheaper transactions help users but may reduce near-term burn.
- Policy debates continue. Issuance, staking concentration, MEV, and validator incentives remain active governance topics.
The strongest interpretation is not “ETH has no cap, so supply does not matter” or “ETH is always deflationary.”
The accurate view is:
Ethereum has a dynamic supply policy where net supply equals issuance minus burns, and the result depends on real network conditions.
Expert tips for interpreting ETH supply data
Use supply direction, not just the headline number
The total count matters, but the trend often matters more.
Ask:
- Is supply rising or falling over the past 7 days?
- Is the burn rate higher than issuance?
- Are blob fees contributing meaningfully to burns?
- Is mainnet activity low because users moved to L2s?
- Is staking participation increasing?
A flat supply number without context is not very useful.
Separate ETH burn from Ethereum revenue
People sometimes treat all fees as “revenue” or all fees as “burn.”
That is sloppy.
Base fees are burned. Priority fees and MEV-related payments are transfers. DEX fees go to liquidity providers. Sequencer fees on L2s may go to different parties. These flows have different economic meanings.
Watch the base fee, not just transaction count
A million cheap transactions may burn less ETH than fewer expensive transactions.
ETH burn is driven by gas used multiplied by base fee, not raw transaction count alone.
Be careful with “deflationary” claims
ETH can be deflationary over one period and inflationary over another.
A headline like “Ethereum is deflationary” should always be paired with a timeframe.
Better wording:
- “ETH supply decreased over the past 30 days.”
- “ETH has been net inflationary since [date].”
- “Burns exceeded issuance during this period.”
Do not count derivatives as ETH
If you are analyzing ETH supply, do not add stETH, WETH, bridged ETH, or exchange IOUs on top of native ETH. That double-counts claims and representations.
Common mistakes about Ethereum coin supply
Mistake 1: Thinking Ethereum has a 21 million cap
That cap belongs to Bitcoin. Ethereum has no fixed maximum supply.
Mistake 2: Assuming no cap means unlimited inflation
Ethereum issuance is protocol-defined and much lower after the Merge. Net supply can rise or fall depending on burns.
Mistake 3: Treating staked ETH as removed from supply
Staked ETH is still ETH. It is not burned.
Mistake 4: Counting WETH as extra ETH
WETH is a wrapped representation of ETH. It does not increase native supply.
Mistake 5: Believing all gas fees are burned
Only the base fee is burned. Priority fees and some MEV-related payments are transfers.
Mistake 6: Using market cap sites without checking the metric
A market cap page may show circulating supply, not protocol total supply. Read the label.
Mistake 7: Ignoring layer 2 effects
Ethereum activity increasingly happens on L2s. That changes how user demand translates into L1 and blob fee burns.
Mistake 8: Quoting an ETH supply number without a timestamp
Because ETH supply changes continuously, a precise number without a timestamp is incomplete.
Key takeaways
- Ethereum has roughly 120 million ETH in existence.
- The exact supply changes continuously as validators receive issuance and fees are burned.
- ETH does not have a fixed maximum supply like Bitcoin.
- EIP-1559 burns the base fee from transactions.
- EIP-4844 introduced blob fees, which can also be burned.
- Staked ETH remains part of total ETH supply.
- WETH, stETH, rETH, cbETH, and bridged ETH are not additional native ETH.
- ETH can be inflationary or deflationary depending on network activity and staking issuance.
- For exact current supply, use a live tracker and record the timestamp and block number.
FAQ
How many Ethereum coins are left to mine?
None.
Ethereum no longer uses mining. Since the Merge, Ethereum has used proof of stake. New ETH is issued to validators, not miners.
Is Ethereum supply unlimited?
Ethereum has no fixed maximum cap, but that does not mean uncontrolled issuance. ETH issuance follows protocol rules, and fee burns can offset or exceed new issuance.
Can Ethereum ever reach 21 million ETH like Bitcoin?
Not under the current design. Ethereum’s supply is around 120 million ETH, far above 21 million. Its monetary model is not designed to converge to Bitcoin’s cap.
Is ETH deflationary right now?
That depends on the current burn rate versus validator issuance. ETH is deflationary during periods when burned fees exceed new issuance. Check a live supply dashboard for the current period.
Why did ETH supply go down after the Merge?
The Merge removed proof-of-work miner rewards and replaced them with lower proof-of-stake issuance. Because EIP-1559 already burned base fees, Ethereum became capable of net supply reduction when burns exceeded issuance.
Does burning ETH make the price go up?
Not automatically. Burns reduce supply, but price also depends on demand, liquidity, macro conditions, leverage, regulation, staking behavior, and broader crypto market sentiment.
Supply reduction can be supportive, but it is not a price guarantee.
Does staking ETH create new ETH?
Validator rewards create new ETH. The act of staking existing ETH does not create supply by itself. Rewards are issued over time for validator participation.
Is stETH counted in Ethereum supply?
No. stETH is not native ETH. It is a liquid staking token representing a claim on staked ETH. Counting both ETH and stETH as separate ETH supply would double-count.
Is WETH counted as Ethereum supply?
WETH is a wrapped ERC-20 representation of ETH. It does not add to total ETH supply.
Does ETH on Arbitrum, Optimism, Base, or another L2 count as Ethereum supply?
Native ETH remains part of Ethereum’s broader accounting, but bridged or represented ETH on L2s should not be counted as newly created ETH. It is usually backed by ETH locked or accounted for through bridge and rollup systems.
Why do CoinGecko and Etherscan show different ETH supply numbers?
They may use different definitions, update frequencies, rounding, or methodologies. One may emphasize circulating supply while another tracks total supply. For precise research, cite the source and timestamp.
What happens if someone sends ETH to the wrong address?
If the ETH is sent to a valid but inaccessible address, it may be permanently lost economically. However, it usually remains counted in total supply unless it is burned through a protocol-recognized burn mechanism.
Are Ethereum gas fees always paid in ETH?
On Ethereum mainnet, gas is paid in ETH. On L2s, users may experience different fee abstractions, but ETH remains central to Ethereum settlement and rollup economics.
What is the best site to check current ETH supply?
For a quick answer, Etherscan and ultrasound.money are commonly used. For market context, CoinGecko is useful. For research-grade work, compare multiple sources and record the timestamp and block number.
Final verdict
Ethereum has about 120 million ETH, but the exact number is not fixed.
The better way to understand ETH supply is not as a countdown to a hard cap. It is a live equation:
ETH supply = existing ETH + validator issuance - burned fees - penalties
That makes Ethereum’s monetary policy more complex than Bitcoin’s, but also more responsive to how the network is actually used.
If you need a simple answer, say: roughly 120 million ETH exist.
If you need an accurate answer, check a live supply tracker at the moment you need the data and record the source, timestamp, and block number.
