# Parallel calls

> Fan one function out over many inputs with .map(), each input in its own container.

Canonical page: https://klio.tech/docs/parallel

## Fan out with .map()

```python
from klio_compute import function

@function(cpu=1, memory="1Gi")
def square(x: int) -> int:
    return x * x

if __name__ == "__main__":
    for result in square.map([1, 2, 3, 4]):
        print(result)
```

`.map()` sends every input to its own container and gives back the results. With `@function`, results arrive as calls finish; a failed input is reported and yields `None`. Each container is billed for the seconds it runs.

## With the Modal-compatible API

```python
import klio_compute.compat as modal

app = modal.App("thumbnails")

@app.function()
def thumbnail(path: str) -> str:
    ...

@app.local_entrypoint()
def main():
    print(list(thumbnail.map(["a.png", "b.png", "c.png"])))
```

Here `.map()` returns results in input order, as Modal does, and keeps a bounded number of calls in flight however long the input is.