What an omnichain build costs for a small dapp
A two-chain omnichain dapp can take 15–25 engineer-days before audit; message design, retries, testing and variable gas shape the real build estimate.
The Top Crypto Editors 3 min read
What should you budget to make a small dapp omnichain? For a two-chain EVM app with one cross-chain action, plan on roughly 15–25 engineer-days before an independent security audit. That estimate assumes an existing dapp and a narrow feature, such as sending a message that triggers an action on the destination chain; a new token bridge or broad multi-chain rewrite is a different project.
The key design choice is what crosses the boundary: a message, a token, or both. An omnichain message usually requires a source-chain transaction, delivery through a messaging layer, and destination-chain execution; the destination call can fail even after the source transaction succeeds. When comparing designs, omnichain is a useful term for the broader pattern, but your estimate should follow the exact route and application behavior you intend to support.
What work goes into an omnichain build?
A minimal two-chain implementation often takes 15–25 engineer-days because the cross-chain state machine needs more than a send function. A practical breakdown is:
- Contract and message design: 4–7 days to define trusted senders, payloads, replay protection, and destination permissions.
- Application integration: 3–5 days to connect the dapp’s user flow, transaction status, and destination execution.
- Testing and failure handling: 5–8 days for delayed, duplicated, rejected, or retried messages and chain-specific configuration.
- Deployment and operations: 3–5 days for deployment scripts, monitoring, alerts, and runbooks.
These are planning ranges, not vendor quotes, and assume one experienced engineer with a working codebase. Parallel work may shorten the calendar time without reducing the total effort.
What makes the estimate rise?
More chains, more message types, token accounting, and recovery paths all add work because each expands the combinations the team must test. A transfer that depends on destination liquidity, for example, needs a policy for what users see when execution is delayed or cannot complete. Treat retries and refunds as product behavior to specify, not edge cases to leave for launch week.
For most small dapps, the better first release is one destination chain and one narrowly defined message flow. It keeps permissions and failure states reviewable while showing whether cross-chain use matters to users. Adding chains later still requires configuration and testing for each route, but does not automatically require redesigning the whole application.
How do you estimate the running cost?
Separate engineering labor from transaction fees and any messaging charges. Each cross-chain action can incur gas on the source and destination, while the delivery layer may quote or charge separately; the amount changes with network conditions, payload, and execution work. Model it as source gas plus delivery fee plus destination gas, then measure representative transactions on the intended chains before setting a user price.
Keep the audit outside the 15–25-day build estimate and scope it after the contracts and message paths are stable. Your next checkpoint is the first end-to-end test on both chains: record source cost, delivery fee, destination cost, and the recovery path when destination execution fails.