◆Fees
What a transaction actually costs
Execution is the small part. Data availability, settlement, and proving are the rest — and a fee model that ignores them is subsidised.
A Layer 2 fee is not just execution. It carries a share of what it costs to publish data and settle on Solana, plus the cost of proving. Any fee model that ignores those is subsidised, and subsidies end.
01Composition
What you are paying for
- Execution
Compute and state access on Solieum. The smallest component at scale.
- Data availability
Publishing the batch so the chain can be verified. Usually the largest component, and the one that scales with usage.
- L1 settlement
Committing state roots to Solana. Fixed per batch, so it shrinks per transaction as batches fill.
- Proving
Generating or standing ready to defend correctness proofs.
Why batching makes small transactions viable
Settlement cost is per batch, not per transaction. One thousand transactions in a batch each carry one thousandth of the settlement cost. This is the entire economic argument for a rollup — and it means per-transaction cost depends on how full batches are, which is why a single headline fee figure is misleading.
Illustrative numbers, chosen round — the mechanism is what matters. One batch pays for one settlement, however many transactions share it.
EXAMPLE — labelled as such
────────────────────────────────────────
Fixed cost to settle one batch on L1: C
Transactions in the batch: N
Settlement share per transaction: C / N
N = 10 → C/10 per transaction
N = 100 → C/100
N = 1,000 → C/1,000
Same guarantee, thousandth of the share.
The data component scales differently —
per byte, not per batch — which is why
fee = execution + data + settlement/N
and why fuller batches mean cheaper actions.◆Asset
Paid in SOL — no token required
Fees are charged in SOL, the asset users already hold and wallets already understand. Solieum is a rollup: its security is Solana's, so there is nothing for a separate gas token to secure, and the operator's own costs on Solana — publishing batches, bonding roots — are paid in SOL too. Charging users in the same asset means no fee oracle and no price exposure between what is charged and what settlement costs.
If a Solieum governance token is ever issued, it governs the economic surface and pays for work — verifying, challenging, staked sequencing — and is never gas and never collateral (ADR-0018). It has no fee-path role: the discount once reserved for it is withdrawn, so SOL stays the rollup’s only fee asset and a user never needs the token to use the chain. Any change to that would be announced with the reasoning, not discovered in a fee schedule.
On the devnet chain today the fee is flat: 5,000 lamports per transaction, charged in SOL to the fee payer. On the beta chain it was burned rather than paid to the sequencer up to block 100 and kept by the operator from that block until the chain stopped at block 158 (ADR-0017), a switch the chain crossed on 2026-09-10 and which each block reports for itself — the same order as a Solana signature fee, and a placeholder for the composition above rather than a market. Underneath it, the measured floor for Solana's data and settlement share is about 280 lamports per transaction on modelled sustained traffic, from the cost model in the repository run on Solana's actual posting mechanic; sparse traffic compresses worse and costs more. And the flat fee does not cover it. Measured from the public endpoint on 2026-09-04: settling one block holds about 4.2 million lamports on Solana, almost all of it rent, of which the 100,000-lamport bond comes back. A block therefore needs about 820 transactions at the flat fee to pay for its own settlement, and the devnet chain is subsidised by its payer by three orders of magnitude. That is fine for a test chain and it is disclosed here; it is not a chain anyone should call self-supporting. Whether the mainnet fee moves with cost, or the subsidy is accepted and disclosed, is a decision still open — not code.
◆Split
Who pays what
| Party | Pays | Detail |
|---|---|---|
| User | the transaction fee, in SOL | Execution + data share + settlement share. The only party paying by default. |
| Application (optional) | sponsored fees | An app can pay on behalf of its users where the account model supports it — a product choice, not a protocol subsidy. |
| Operator | infrastructure + any spike absorption | If the operator chooses to absorb an L1 fee spike, that is a disclosed, bounded decision — the spike policy above — never a silent loss building up. |
| Nobody | hidden spread | The commitment in the principles: no undisclosed margin between what is charged and what settlement actually costs. |
02Volatility
When Solana fees spike
Base-layer fees are not constant. When they rise sharply, an L2 has three options and must pick one in advance rather than during the incident.
Pass through
User fees rise with L1. Honest and sustainable; users feel the volatility.
Absorb
The operator runs at a loss to keep fees stable. Fine briefly, and a hidden liability if it becomes the plan.
Throttle
Slow batch posting until costs normalise. Protects economics at the cost of settlement latency.
Solieum will publish which of these it does, and the threshold at which behaviour changes, before mainnet. An undefined policy here becomes an improvised decision during exactly the wrong week.
What we commit to
- Fees are charged in SOL. No token is required to transact, and the governance token, if ever issued, has no fee-path role: ADR-0018 withdrew the discount once reserved for it.
- Fees track real cost — no figure that only holds while someone is subsidising it. Today's flat fee does not yet, and the measured gap is stated below rather than hidden.
- The cost model is published, so a quoted fee can be checked rather than trusted.
- Fee changes are announced before they take effect.
- No hidden spread between what is charged and what is paid to Solana.
Building something the base layer can't hold?
Tell us the workload. If Solieum is the wrong answer for it, we would rather say so early than have you find out at launch.