Swiss Voice & Speech API • v1.8 • ISO 27001 Certified
The Swiss Speech API for Multilingual Voice AI & Swiss Dialects
High-accuracy Speech-to-Text (STT) and natural Text-to-Speech (TTS) in English, German, French and Italian – including Swiss dialects. Available as ultra-low-latency real-time WebSocket streams or high-throughput batch REST APIs.
Suisse Speech
More information on our dedicated website
Swiss speech-to-text & text-to-speech in all national languages, English and Swiss dialects: documentation, pricing, self-service sign-up. 60 free minutes, billed per second.
🇨🇭 Hosted in Zurich
API servers in a Zurich data centre. Full compliance with the Swiss revFADP and EU GDPR.
⚡ < 500 ms TTFA
Sub-500ms time-to-first-audio for seamless conversational AI voicebots.
🔒 ISO/IEC 27001
ISO/IEC 27001:2022 certified architecture for strict enterprise compliance.
⏱️ Second-Accurate Billing
No minimum call fees or lock-ins. Pay strictly for the exact audio duration consumed.
TTS & STT: Enterprise Speech Processing for Developers
Natural voices and high-accuracy transcription in English, German, French, Italian and Swiss dialects – via robust REST endpoints and duplex WebSockets.
POST /v1/tts & /v1/tts/stream
Text-to-Speech (TTS)
Lifelike voices in English, German, French and Italian with authentic Swiss dialect intonation. Engineered for automated telephony, conversational AI, and voice applications.
- Synthesis in English, German, French, Italian and Swiss dialects (Zurich, Bern, Basel, St. Gallen, Alpine)
- Real-time WebSocket streaming (<500ms time-to-first-audio chunk)
- Flexible audio encoding: MP3, Opus, WAV, and 16-48kHz Linear PCM
- Granular speed and tone controls with expressive native voices (Anna, Stefan)
POST /v1/stt & /v1/stt/stream
Speech-to-Text (STT)
High transcription accuracy across English, German, French, Italian and all Swiss regional dialects, robust against background noise and accents.
- Multilingual transcription plus direct Swiss dialect-to-standard or dialect text
- Real-time WebSocket streaming with interim results and confidence scores
- High-throughput asynchronous batch processing for call centres & meetings
- Word-level timestamps, speaker diarisation, and custom vocabulary support
Engineered by Developers, for Developers
Clean REST endpoints, standard WebSockets, SDKs, and a free synthetic sandbox mode for CI/CD test automation.
Integrated in 5 Minutes
Incorporate multilingual Swiss voice and dialect capabilities into your application with a single HTTP request or WebSocket stream. Compatible with all programming languages and telephony stacks.
API Specifications
| Base URL | https://api.suisse-speech.ch/v1 |
| Auth Header | X-API-Key: sv_live_... |
| Streaming WebSocket | wss://api.suisse-speech.ch/v1/tts/stream |
| Audio Formats | mp3, wav, opus, pcm (16-48kHz) |
| Interactive Docs | OpenAPI 3.0 (Swagger UI) ↗ |
Free Synthetic Sandbox Mode for CI/CD
Using a sandbox API key (sv_test_...) with header X-Suisse-Sandbox-Mode: synthetic provides structurally accurate audio and transcripts without consuming quota — perfect for CI/CD test suites.
curl -X POST https://api.suisse-speech.ch/v1/tts \
-H "X-API-Key: $SUISSE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"text": "Good morning! How can I help you today?",
"voice": "anna",
"language": "en-CH",
"speed": 1.0,
"format": "mp3"
}' \
--output greeting.mp3import requests
url = "https://api.suisse-speech.ch/v1/stt"
headers = {"X-API-Key": "sv_live_..."}
with open("customer_call.wav", "rb") as f:
files = {"file": ("audio.wav", f, "audio/wav")}
data = {"language": "en-CH", "dialect": "auto"}
res = requests.post(url, headers=headers, files=files, data=data)
print(res.json()["transcript"])All Swiss Languages & Regional Dialects Built In
Alongside English, standard German, Swiss French and Swiss Italian, our AI natively masters regional vocabularies, phonetics, and speech patterns across all Swiss dialects.
Zurich & Eastern Switzerland
Accurate modelling of Zurich, Eastern Switzerland, and Greater Metropolitan speech patterns.
Bern & Western Cantons
Authentic recognition and synthesis of Bernese, Basel, and Western cantonal dialects.
Central Switzerland, Valais & Graubünden
High transcription accuracy even on complex alpine dialects from Lucerne, Valais, and Graubünden.
Enterprise Performance & Reliability
Measured performance on Swiss high-performance infrastructure for mission-critical enterprise workloads.
Time-to-First-Audio
< 480 ms
Streaming latency for interactive telephony
STT dialect accuracy
97.4%
Dialect recognition benchmark accuracy
Service SLA
99.95%
Availability commitment for enterprise customers
Information security
ISO 27001
Certified ISMS, Swiss revFADP & EU GDPR
Frequently Asked Questions (FAQ)
Which languages and Swiss dialects are supported by the Speech API?
The API supports English, German, French and Italian, as well as all major Swiss dialect groups – including Zurich, Bern, Basel, St. Gallen, Valais, Graubünden and Central Switzerland, plus Swiss French and Swiss Italian varieties. Languages and dialects can be specified explicitly or detected automatically.
How low is the latency of the real-time WebSocket streaming API?
Streaming delivers the first audio chunk in under 500 ms (Time-to-First-Audio), allowing natural, interruption-free phone conversations with conversational AI bots.
Is my audio data used for training AI models?
No. All customer audio and transcripts remain strictly confidential, are never used for model training, and are not stored. Our API servers are located in Zurich; where each processing step takes place is set out in the Suisse Speech privacy policy (opens in a new tab).
How can I test the Speech API for free?
Sign up for free on suisse-speech.ch to receive 60 free minutes of starting credit for Speech-to-Text and Text-to-Speech – no credit card required and no expiry. In addition, an unlimited synthetic sandbox environment is available free of charge for automated CI/CD testing.
Ready to Integrate Swiss Voice AI & Dialects into Your App?
Get started today with 60 free minutes on suisse-speech.ch – no credit card required – or contact our Swiss engineering team to discuss enterprise custom requirements.