> ## Documentation Index
> Fetch the complete documentation index at: https://docs.asterwise.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Responses API

> Use the OpenAI Responses-shaped endpoint at /v1/responses.

## Endpoint

Set model to aster-code or aster-work. Send input as text or as supported structured input. Streaming and tool continuation depend on the exact client and route combination you use.

```bash theme={null}
POST https://api.asterwise.dev/v1/responses
Authorization: Bearer $ASTER_API_KEY
Content-Type: application/json
```

## Minimal request

```bash theme={null}
curl https://api.asterwise.dev/v1/responses \
  -H "Authorization: Bearer $ASTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"aster-code","input":"Return the smallest safe patch plan."}'
```

## Continuation and tools

* Return tool outputs with the identifiers supplied by the preceding response.
* Keep the model name stable within a continuation unless your harness deliberately starts a new request.
* Do not assume every OpenAI Responses field is supported; begin with the fields your documented harness setup needs.

## Next steps

* [Handle errors and retries](/errors-fallback)
* [Check supported capabilities](/products-capabilities)