Multilingual Support

Multi-language support for ambient sessions and clinical documentation

Quick Summary

Multilingual support lets patients and providers speak in their preferred language during clinical conversations, while Suki automatically generates the final clinical note in English.

Last updated:December 2025

Multilingual support is supported for Mobile SDK and APIs only. The Web SDK and Headless Web SDK do not currently support multilingual sessions.

Overview

Multilingual support lets patients and providers speak in their preferred language during clinical conversations, while Suki automatically generates the final clinical note in English. This removes the need for manual translation and makes healthcare more accessible to diverse patient populations.

When you enable multilingual support for an ambient session, you get the following benefits:

• Patient comfort: Patients can communicate in their native language, leading to more accurate information sharing
• Better care quality: When patients speak in their preferred language, they provide more detailed and accurate information
• No translation needed: Clinicians don’t need to translate conversations manually, Suki handles it automatically
• EHR compatibility: All notes are generated in English, ensuring compatibility with standard EHR systems
• Wider accessibility: Support for 80+ languages makes healthcare more inclusive

How It Works

When you enable multilingual support for an ambient session, Suki automatically:

1. Detects the language spoken during the conversation
2. Transcribes the audio in the detected language
3. Translates and processes the conversation content
4. Generates the clinical note in English

The transcript API returns a lang_id field that identifies which language was detected for each segment of the conversation. This helps you understand what language was spoken during different parts of the session.

How To Enable Multilingual Support

Enable multilingual support when creating an ambient session. The exact method depends on whether you’re using the Mobile SDK or APIs.

Mobile SDK Example:

Swift

let sessionContext = [
    SukiAmbientConstant.kSessionId: encounterId,
    SukiAmbientConstant.kMultilingual: true // Enable multilingual support
] as [String : AnyHashable]

SukiAmbientCoreManager.shared.createSession(
    with: sessionContext,
    onCompletion: { result in
        switch result {
        case .success(let sessionResponse):
            let sessionId = sessionResponse.sessionId
            // Store sessionId for future use
        case .failure(let error):
            // Handle error
        }
    }
)

API Example:

Python

payload = {
    "ambient_session_id": "123dfg-456dfg-789dfg-012dfg",  # Optional
    "encounter_id": "123dfg-456dfg-789dfg-012dfg",  # Optional
    "multilingual": True  # Enable multilingual support
}

response = requests.post(
    "https://sdp.suki.ai/api/v1/ambient/session/create",
    json=payload,
    headers={"sdp_suki_token": "<sdp_suki_token>"}
)

For complete API documentation, see the Create Ambient Session API.

Multilingual support is disabled by default.

You must explicitly enable it when creating a session. Once enabled, it applies to the entire session and cannot be changed mid-session.

Supported Languages

Suki supports over 80 languages for ambient sessions. The following languages are currently supported:

Spanish	Norwegian	Macedonian	Kazakh	Yoruba
Italian	Finnish	Hungarian	Icelandic	Telugu
English	Vietnamese	Tamil	Marathi	Khmer
Portuguese	Thai	Hindi	Maori	Malayalam
German	Slovak	Estonian	Swahili	Lao
Japanese	Greek	Urdu	Armenian	Punjabi
Polish	Czech	Latvian	Belarusian	Gujarati
Russian	Croatian	Slovenian	Nepali	Somali
Dutch	Danish	Azerbaijani	Occitan	Bengali
Indonesian	Tagalog	Hebrew	Lingala	Georgian
Catalan	Korean	Lithuanian	Maltese	Assamese
French	Romanian	Persian	Tajik	Mongolian
Turkish	Bulgarian	Welsh	Luxembourgish	Myanmar
Swedish	Galician	Serbian	Hausa	Shona
Ukrainian	Bosnian	Afrikaans	Uzbek	Amharic
Malay	Arabic	Kannada	Pashto	Sindhi
Chinese (Cantonese)	Chinese (Mandarin)

Language Code Reference

Use the following table to map each lang_id to its corresponding language. This helps you interpret the language codes returned by the Transcript API.

We regularly update this list as we add support for new languages. If a language is not included in this table, it is not yet supported.

A

Language	Language ID (lang_id)
afrikaans	af
amharic	am
arabic	ar
armenian	hy
assamese	as
azerbaijani	az

B

Language	Language ID (lang_id)
belarusian	be
bengali	bn
bosnian	bs
bulgarian	bg

C

Language	Language ID (lang_id)
catalan	ca
chinese	zh
chinese_cantonese	yue
chinese_mandarin	cmn
croatian	hr
czech	cs

D

Language	Language ID (lang_id)
danish	da
dutch	nl

E

Language	Language ID (lang_id)
english	en
estonian	et

F

Language	Language ID (lang_id)
finnish	fi
french	fr

G

Language	Language ID (lang_id)
galician	gl
georgian	ka
german	de
greek	el
gujarati	gu

H

Language	Language ID (lang_id)
hausa	ha
hebrew	he
hindi	hi
hungarian	hu

I

Language	Language ID (lang_id)
icelandic	is
indonesian	id
italian	it

J

Language	Language ID (lang_id)
japanese	ja

K

Language	Language ID (lang_id)
kannada	kn
kazakh	kk
khmer	km
korean	ko

L

Language	Language ID (lang_id)
lao	lo
latvian	lv
lingala	ln
lithuanian	lt
luxembourgish	lb

M

Language	Language ID (lang_id)
macedonian	mk
malay	ms
malayalam	ml
maltese	mt
maori	mi
marathi	mr
mongolian	mn
myanmar	my

N

Language	Language ID (lang_id)
nepali	ne
norwegian	no

O

Language	Language ID (lang_id)
occitan	oc

P

Language	Language ID (lang_id)
pashto	ps
persian	fa
polish	pl
portuguese	pt
punjabi	pa

R

Language	Language ID (lang_id)
romanian	ro
russian	ru

S

Language	Language ID (lang_id)
serbian	sr
shona	sn
sindhi	sd
slovak	sk
slovenian	sl
somali	so
spanish	es
swahili	sw
swedish	sv

T

Language	Language ID (lang_id)
tagalog	tl
tajik	tg
tamil	ta
telugu	te
thai	th
turkish	tr

U

Language	Language ID (lang_id)
ukrainian	uk
urdu	ur
uzbek	uz

V

Language	Language ID (lang_id)
vietnamese	vi

W

Language	Language ID (lang_id)
welsh	cy

Y

Language	Language ID (lang_id)
yoruba	yo

Best Practices

• Enable when needed: Only enable multilingual support when you expect conversations in multiple languages. This optimizes performance.
• Set patient language preference: If you know the patient’s preferred language, you can display this information in your UI to help providers prepare.
• Monitor language detection: Use the lang_id from transcripts to understand language usage patterns in your application.
• Test with your languages: Verify multilingual support works correctly with the languages your patients commonly use.
• Note language in UI: Consider displaying the detected language in your UI so providers know what language was spoken.

Related APIs

Use these APIs to work with multilingual support:

Get Session Transcript

Retrieve transcripts with language detection information (lang_id)

Create Session

Create ambient sessions with multilingual support enabled