Key findings
- Execution gas is only one component of the user’s total cost.
- Compare the entire route from starting assets to the desired final state.
- Additional chains can improve access while increasing liquidity and operational fragmentation.
The cheapest transaction may not be the cheapest experience
Low execution fees directly benefit a user who already holds the required asset on the chain. A user arriving from elsewhere also faces transfer costs, waiting, liquidity and recovery steps. A single transaction quote covers only part of the job.
Specify starting and ending asset states. Official bridges, liquidity routes and exchange routes involve different counterparties and timing even for the same chain pair. The chain alone does not determine total cost.
The fee layers visible in OP Stack
The Isthmus execution specification describes an operator fee alongside execution and L1 data costs. It is a concrete reason not to interpret all costs as execution gas, not evidence that every L2 has the same fee structure. [1]
Pectra’s announced blob-capacity changes also affect the conditions for data publication. More capacity does not imply an equal percentage reduction in every user fee; demand, pricing policies and transaction types matter. [2]
| Layer | Observation unit | Product decision |
|---|---|---|
| Execution | Actual fee per task | Instructions and batching |
| Data publication | Chain-specific charging | Payload size and aggregation |
| Operator | Additional policy-based costs | Supported chains and budget |
| Asset movement | Route cost and waiting | Assets and deposit/withdrawal flow |
| Operations | Support, settlement and inventory | Number of routes supported |
CBK analysis. This is an analytical framework, not observed customer performance or a market benchmark.
Fragmented liquidity can move costs elsewhere
More chains can bring services closer to customers while requiring more inventory and states to manage. Thin liquidity can worsen execution prices even when displayed fees are low. Measure price impact separately from network charges.
With hypothetical per-task costs of 0.10 on one chain and 0.02 on another, a transfer costing 2 requires 25 subsequent tasks to offset the direct difference: 2 ÷ (0.10 − 0.02) = 25. Support and price risk are excluded, so this is a minimum-condition illustration, not a fee forecast.
| Use pattern | Cost to inspect first |
|---|---|
| One-off task | Initial asset transfer and explanation |
| Repeated use | Per-task cost and failure rate |
| Institutional operation | Settlement, recovery and controls |
| Multichain service | Liquidity inventory and route operations |
CBK analysis. This is an analytical framework, not observed customer performance or a market benchmark.
Low prices and predictable prices are different benefits
An occasional consumer and a business processing repeated tasks need not value cost stability equally. Companies may need tail-cost estimates and the frequency of budget exhaustion alongside average fees.
A flat charge across routes can cross-subsidize expensive paths; passing through every cost can complicate the experience. This is a pricing decision as well as a network choice. Connect route-level costs to the customer-facing price.
| Policy | Benefit | Cost to examine |
|---|---|---|
| Fixed charge | Predictable user price | Spikes and cross-subsidy |
| Pass-through | Reflects actual cost | Comprehension, quotes and abandonment |
| Partial sponsorship | Less initial friction | Budget, abuse and post-subsidy use |
CBK analysis. This is an analytical framework, not observed customer performance or a market benchmark.
Test the round trip before expanding
Run the same task from the same starting assets and include recovery of funds afterward. Test insufficient balances, route interruption and a mistaken network selection. These reveal costs absent from the advertised quote.
Multichain support can be valuable when customers are already distributed across networks. Expansion should follow their starting points and recurring tasks, rather than chain counts or fee slogans. Additional connections also warrant operational-risk review. [3]
Scope & limitations
A specific OP Stack specification is used as an example, not a model of every L2. Bridge security is not ranked. Break-even figures use hypothetical costs.
CryptoBridge Korea analysis based on public materials. Source facts are distinguished from interpretation and hypothetical calculations. References to organizations or projects do not imply their sponsorship or review of this paper.
Sources & scope
- OP Stack specification, Isthmus execution engine ↗Isthmus fee specification; not representative of every Layer 2.
- Ethereum Foundation, Pectra Mainnet Announcement ↗Announcement of upgrade features and the 7 May 2025 activation schedule.
- FSB, The Financial Stability Implications of Tokenisation ↗Analysis of tokenization and financial stability, not a safety certification for any product.