# `LlmCall.Response`
[🔗](https://github.com/gilbertwong96/llm_call/blob/main/lib/llm_call/response.ex#L1)

A provider-neutral completion response.

`content` is a list of content parts, never provider-specific choices or
output items. `raw` is a provider snapshot, retained only when
`LlmCall.Config.capture_raw` asks for it, and is never a second response
model.

# `t`

```elixir
@type t() :: %LlmCall.Response{
  content: [LlmCall.Content.part()],
  id: String.t() | nil,
  model: String.t() | nil,
  raw: map() | nil,
  stop_reason: LlmCall.stop_reason(),
  usage: LlmCall.Usage.t() | nil
}
```

# `text`

```elixir
@spec text(t()) :: String.t()
```

All text parts joined in order.

# `tool_calls`

```elixir
@spec tool_calls(t()) :: [LlmCall.ToolCall.t()]
```

The tool calls in this response, in order.

---

*Consult [api-reference.md](api-reference.md) for complete listing*
