Why Most "Swiss German" TTS Is Not Swiss German at All
Most widely marketed "Swiss German text to speech" tools actually generate Swiss Standard German – High German read with a Swiss accent – and not Swiss German dialect. The difference matters: Swiss German (Schwiizerdütsch) is a family of spoken Alemannic dialects used in everyday life across German-speaking Switzerland, while Swiss Standard German is the written language of newspapers, contracts and school essays. The two differ in vocabulary, grammar and sound. A person from Zurich says "Ich gang hei", not "Ich gehe nach Hause".
As of July 2026, Microsoft Azure's de-CH voices Leni and Jan speak Swiss Standard German. Narakeet, SpeechGen, PlayHT, ttsfree and FineVoice resell these same Azure voices under the label "Swiss German". ElevenLabs has no Swiss German dialect voices either; its catalogue offers "Standard German with Swiss Accent". If you need audio that sounds the way people in Zurich, Basel or the Valais actually talk, an accent model cannot deliver that.
schwiizerdütsch.com takes the other route: an own AI voice model, trained from scratch on Swiss German – not a relabelled High German model – and co-developed on public research data such as SwissDial from ETH Zürich.
How It Works: Type High German, Get Dialect Audio
You type High German – or any Swiss German spelling you like – and the model returns spoken Swiss German dialect audio. This works because Swiss German has no standardised orthography: "hüt", "hütt" and "hüüt" can all mean "today", depending on who writes it. A built-in text normalisation layer converts High German or any Swiss spelling automatically into the correct dialect pronunciation. Typing standard German in and getting dialect audio out is a feature no accent-based voice can offer.
- Step 1: Paste your text – High German or any Swiss German spelling.
- Step 2: Pick a voice and, if you want, adjust the speaking speed between 0.5× and 2×.
- Step 3: Generate and download the audio as MP3, WAV, OGG, FLAC or PCM.
Voices and Dialects
Five voices cover four dialect regions: Zurich, Valais, St. Gallen and Basel. Marco can be generated directly in the Studio today; the four other voices can be previewed with audio samples. The model behind them was trained from scratch on Swiss German speech – not a High German voice with a filter on top.
| Voice | Dialect | Availability |
|---|---|---|
| Marco | Züridütsch (Zurich German) | Generate in the Studio |
| Toni | Walliserdütsch (Valais German) | Audio sample |
| Nicole | St. Gallerdütsch (St. Gallen German) | Audio sample |
| Robert | Baseldütsch (Basel German) | Audio sample |
| Lewin | Züridütsch (Zurich German) | Audio sample |
Probiers selber us: Free and Without Signup
The demo on the homepage works without any signup – paste a sentence and hear Swiss German dialect. With a free account you additionally unlock the Studio and can generate up to 50,000 characters per day. The current release is non-commercial; if you want to use the audio in a commercial project, a short mail is all it takes to start the conversation.
schwiizerdütsch.com is a Swiss company and operates under the Swiss data protection law (nDSG).
API and Python SDK for Developers
A REST API with Bearer-token authentication and a Python SDK – installed with pip install schwiizerduetsch – let you generate Swiss German dialect audio from your own application. The relevant numbers:
- Output formats: MP3, WAV, OGG, FLAC and PCM
- Speaking speed: adjustable from 0.5× to 2×
- Request size: up to 10,000 characters per request
- Batch processing: generate multiple texts in one job
- Input: High German or any Swiss German spelling – normalisation is handled server-side
To be transparent about the limits: streaming, WebSocket connections and SSML are not offered at the moment. What is listed here is what actually works today.
Swiss German Voice Generators Compared (July 2026)
As of July 2026, real Swiss German dialect TTS comes from a small number of Swiss providers – most international tools labelled "Swiss German" output Swiss Standard German instead. The overview:
| Provider | Real dialect? | Voices / dialects | Notes (as of July 2026) |
|---|---|---|---|
| schwiizerdütsch.com | Yes | 5 voices (1 generatable in the Studio, 4 with audio samples) across 4 dialect regions | Free demo without signup, 50,000 characters/day with free account, REST API + Python SDK |
| Plapperi (plapperi.ch) | Yes | 3 Zurich German demo voices (Alba, Brisa, Calder) | Launched January 2026; voices described as placeholders on their own blog; no public pricing page; REST API + Python SDK; streaming listed as coming soon |
| Microsoft Azure (de-CH) | No – Swiss Standard German | 2 voices (Leni, Jan) | Accent, not dialect; the basis for many resellers |
| Narakeet, SpeechGen, PlayHT, ttsfree, FineVoice | No | Resell the Azure de-CH voices | Labelled "Swiss German", output is Swiss Standard German |
| ElevenLabs | No | "Standard German with Swiss Accent" only | No Swiss German dialect voices |
| FHNW research demo (stt4sg.fhnw.ch/tts) | Yes | Research prototype (Gradio) | Free, built for research, not a product |
| suisse-ansage.ch | Yes (niche) | 5 dialects for phone announcements/IVR | Credit packages from CHF 9.90 to CHF 69.90 |
| SlowSoft GmbH | Historical provider | Swiss German and Romansh | Has ceased business operations |
A fair word on Plapperi: the team behind it (Noxenum GmbH, Aarau) works with SwissNLP, its founders come from FHNW research (one is a co-author of the STT4SG corpus), and their system also accepts High German input. That is a serious research background. Today, its public voice line-up consists of three Zurich German voices its own blog calls placeholders, and its stated business model is custom voice finetuning for organisations.
When Another Provider Is the Better Choice
For custom organisational voices, Swiss Standard German or telephone systems, another provider can be the better fit – here is a straight assessment:
- Plapperi is the better choice if you want a custom voice finetuned for your organisation – that is their stated business model – and you value the team's research lineage (ex-FHNW, SwissNLP collaboration).
- Azure and its resellers (Narakeet, SpeechGen, PlayHT and others) are the better choice if you genuinely need Swiss Standard German – for example formal texts that Swiss listeners expect in Standard German anyway; commercial use is governed by their respective terms.
- suisse-ansage.ch is the better choice for telephone announcements and IVR menus with fixed credit packages (CHF 9.90–69.90) across 5 dialects.
- The FHNW demo is the better choice for free research experiments without any product expectations.
Disclaimer
All product and company names mentioned on this page are trademarks of their respective owners. schwiizerdütsch.com is not affiliated with any of the providers mentioned. Competitor information was verified in July 2026 and may change; if you spot an error, write to andrin@iten.to and we will correct it.
Frequently asked questions
What is the difference between Swiss German and Swiss Standard German?
Swiss German is a family of spoken Alemannic dialects used in everyday life across German-speaking Switzerland. Swiss Standard German is the written High German used in newspapers and official documents, pronounced with a Swiss accent. They differ in vocabulary, grammar and sound – which is why accent-based TTS voices do not produce real dialect.
Is there a free Swiss German text to speech tool?
Yes. The demo on the schwiizerdütsch.com homepage works without any signup. With a free account you get access to the Studio and can generate up to 50,000 characters of Swiss German audio per day. The current release is non-commercial; for commercial projects, a short mail is enough to start the conversation.
Can I type High German and get Swiss German audio?
Yes, that is the core feature. The text normalisation accepts High German or any Swiss German spelling and converts it automatically into the correct dialect pronunciation. You do not need to know how to write dialect – paste a standard German sentence, pick a voice and the model speaks it as Swiss German.
Which Swiss German dialects are available?
Marco (Zurich German) can be generated directly in the Studio. Four further voices have audio samples: Toni speaks Valais German, Nicole St. Gallen German, Robert Basel German and Lewin Zurich German. The underlying model was trained from scratch on Swiss German speech, not adapted from a High German model with an accent filter.
Is there a Swiss German text to speech API?
Yes. schwiizerdütsch.com offers a REST API with Bearer-token authentication and a Python SDK (pip install schwiizerduetsch). It supports MP3, WAV, OGG, FLAC and PCM output, playback speed from 0.5 to 2 times, batch processing and up to 10,000 characters per request. Streaming and SSML are currently not offered.
Does ElevenLabs or Azure offer real Swiss German dialect voices?
No, as of July 2026. Microsoft Azure's de-CH voices Leni and Jan speak Swiss Standard German – High German with a Swiss accent. ElevenLabs lists only "Standard German with Swiss Accent" and has no dialect voices. Tools like Narakeet, SpeechGen, PlayHT and FineVoice resell the Azure voices under the "Swiss German" label.
How is schwiizerdütsch.com different from Plapperi?
Both generate real Swiss German dialect from High German input. Plapperi (launched January 2026, in collaboration with SwissNLP, founders from FHNW research) currently offers three Zurich German demo voices, with no public pricing page. schwiizerdütsch.com covers four dialect regions with five voices (Marco generatable in the Studio, four with audio samples), a signup-free demo and 50,000 free characters per day with a free account (as of July 2026).
Can I use the generated audio commercially?
Not under the current release – it is non-commercial, which means the generated audio may not be used in commercial projects yet. If you have a commercial use case, send a mail and describe it; commercial requests are handled individually. Private, research and testing use within the daily character limit is free.