# Language support

LawyerGPT is designed to assist users across multiple legal jurisdictions and languages. Our models are trained to interpret and generate legal content in various languages and styles while adhering to jurisdictional nuances where specified.

## ✅ Supported Languages ( beta )

| Language Name | Code | Notes                  |
| ------------- | ---- | ---------------------- |
| English       | `en` | Full support           |
| Spanish       | `es` | Full support           |
| French        | `fr` | High accuracy          |
| German        | `de` | High accuracy          |
| Polish        | `pl` | Full support           |
| Italian       | `it` | Partial legal coverage |
| Portuguese    | `pt` | Partial legal coverage |
| Dutch         | `nl` | Experimental           |
| Hindi         | `hi` | Limited legal domain   |
| Arabic        | `ar` | Limited legal domain   |

We are actively expanding to support more languages based on demand and available legal datasets.

## Input vs Output Language

You may specify the language for:

* **Input interpretation**: The language in which the question or document is written.
* **Output generation**: The language used to return the response or document.

Use the `language` parameter in your API requests to define your preference. The default is `"en"` (English).

```json
{
  "question": "Jakie są moje prawa jako pracownik?",
  "language": "pl"
}
```

## Regional Formatting

In addition to language, regional terminology, spellings, and legal constructs can be tailored using the `regional_format` parameter:

| Region Code | Format                              |
| ----------- | ----------------------------------- |
| `US`        | American English, U.S. legal system |
| `UK`        | British English, UK laws            |
| `EU`        | Neutral European law alignment      |
| `CA`        | Canadian law alignment              |
| `IN`        | Indian jurisdictional format        |

## Legal Language Quality

* Responses are optimized for **clarity**, **legal precision**, and **reader comprehension**.
* When targeting non-lawyers (e.g., clients), responses are automatically simplified unless specified with `target_reader: "lawyer"`.

## Upcoming Language Support

We are planning to add support for the following languages soon:

* Japanese (`ja`)
* Korean (`ko`)
* Romanian (`ro`)
* Ukrainian (`uk`)
* Turkish (`tr`)

To request prioritization of a language, please contact us at `support@lawyergptai.cc`.