Squire supports multilingual consultations out of the box, letting you configure both the language spoken during a consultation and the language in which the report is generated. Pass language codes when initializing the SDK and when starting a recording to control this behavior.
The inputLanguage parameter passed to startRecording() defines the spoken language Squire listens for during a consultation. Use auto to let Squire detect the language automatically, or pass a specific code for more consistent results.
| Language | Code |
|---|
| Auto-detect | auto |
| Dutch | nl |
| English | en |
| French | fr |
| German | de |
For shorter consultations, pass the specific language code instead of auto. Auto-detection works best when there is enough spoken audio to identify the language reliably — using an explicit code improves accuracy when consultation length is unpredictable.
Output languages
The outputLanguage parameter passed to new Squire({ outputLanguage }) defines the language in which Squire generates the clinical report. This is independent of the input language, so you can transcribe a Dutch consultation and produce an English report.
| Language | Code |
|---|
| Dutch | nl |
| English | en |
| French | fr |
| German | de |
Usage example
The following example initializes the SDK with English as the report language and starts a recording with French as the spoken language.
import Squire from '@squirehealth/sdk';
const squire = new Squire({
outputLanguage: 'en',
});
await squire.startRecording({
inputLanguage: 'fr',
});
locale option in the widget configuration to match the desired interface and output language. Use the same language codes listed in the output languages table above. 