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

# Output formats

Request one or more formats with the `outputTypes` field of `POST /parse`, repeating the field for each. Every request must name at least one, and each applies to every file in the request.

```bash theme={null}
curl -F files=@invoice.pdf -F outputTypes=doclang -F outputTypes=json \
  http://localhost:8080/parse
```

| `outputTypes` | Best for | Structure | Positions |
| - | - | - | - |
| `doclang` | LLM prompts, RAG, agents | Yes | Yes |
| `json` | Application code, pipelines | Yes | Yes |
| `xml` | Applications built on FineReader Engine's XML export | Yes | Yes |
| `txt` | Search indexing, plain extraction | No | No |

Each result is named after its upload with the format as its extension, under the job's ID: `invoice.pdf` requested as `doclang` and `json` produces `{jobId}/invoice.doclang` and `{jobId}/invoice.json`. A split PDF still produces one file per format.

* **DocLang** is an open, AI-native document format and the one we recommend for LLM input. See [DocLang](https://docs.abbyy.com/fine-parser/concepts/doclang.md).
* **JSON** contains pages, text, tables, and layout elements with their positions.
* **XML** is FineReader Engine's XML export. Its options are set in the `[XMLExportParams]` section of the [profile](https://docs.abbyy.com/fine-parser/enterprise/basics/recognition-settings.md).
* **Plain text** is the recognized text in reading order.