Executing Your Programs

The leo execute command executes the Leo program and outputs a transaction object

leo run vs leo execute

Leo offers two commands for running an entry function, and they serve different purposes:

• leo run evaluates the off-chain (circuit) part of the function and prints its outputs locally. It does not generate a zero-knowledge proof, does not produce a transaction, and nothing touches the network. Crucially, it does not run the on-chain final block, so it cannot exercise any logic that reads or writes program state. This makes it fast and ideal for iterating on the circuit logic.
• leo execute compiles the program, synthesizes the circuit, generates an execution proof, and produces a broadcastable Transaction. Pass --broadcast to send that transaction to the network.

For a function without a final block, a typical development loop is to iterate with leo run until it behaves as expected, then switch to leo execute to produce the proof and transaction. If the function has a final block, keep in mind that leo run only covers the off-chain circuit portion. The final block runs on-chain only once a transaction is broadcast and confirmed, so to exercise that logic you need to broadcast (for example, leo execute --broadcast against a local leo devnode).

Running leo run

To evaluate a function locally and see its outputs without producing a proof or transaction, use leo run:

leo run <FUNCTION_NAME> <INPUT_1> <INPUT_2> ...

It compiles the program, evaluates the function, and prints its outputs. Because it only evaluates the circuit (no proof is synthesized), there is no cost summary or transaction:

console output:

       Leo     ... statements before dead code elimination.
       Leo     ... statements after dead code elimination.
       Leo ✅ Compiled 'hello.aleo' into Aleo instructions.

➡️  Outputs

 • <OUTPUT_1>
 • <OUTPUT_2>
 ...

See the leo run reference for the full list of flags.

Running leo execute

leo execute <FUNCTION_NAME> <INPUT_1> <INPUT_2> ...

Optionally, you can execute a function in a remote Leo program by using

leo execute <PROGRAM_NAME>.aleo::<FUNCTION_NAME> <INPUT_1> <INPUT_2> ...

If executing a function from a local program, the leo execute command will first build/compile that program:

console output:

       Leo     2 statements before dead code elimination.
       Leo     2 statements after dead code elimination.
       Leo ✅ Compiled 'hello.aleo' into Aleo instructions.

It will then print out the summary of the execution plan with

🚀 Execution Plan Summary
──────────────────────────────────────────────
🔧 Configuration:
  Private Key:        APrivateKey1zkp...
  Address:            aleo1...
  Endpoint:           https://api.explorer.provable.com/v1
  Network:            <testnet | mainnet>
  Consensus Version:  9

🎯 Execution Target:
  Program:        <PROGRAM_NAME>
  Function:       <FUNCTION_NAME>
  Source:         <local | remote>

💸 Fee Info:
  Priority Fee:   0 μcredits
  Fee Record:     no (public fee)

⚙️ Actions:
  - Transaction will NOT be printed to the console.
  - Transaction will NOT be saved to a file.
  - Transaction will NOT be broadcast to the network.

Finally, an execution cost breakdown will be printed alongside any outputs from the function itself.

📊 Execution Cost Summary for <PROGRAM_NAME>
──────────────────────────────────────────────
💰 Cost Breakdown (credits)
  Transaction Storage:  0.001316
  On-chain Execution:   0.000000
  Priority Fee:         0.000000
  Total Fee:            0.001316
──────────────────────────────────────────────

➡️  Output

  • <OUTPUT_1>
  • <OUTPUT_2>
  ...

Under the hood, leo execute produces a JSON object. This is a Transaction that can be broadcast to the Aleo network. You can view this JSON by passing the --print flag to leo execute.