Microsoft Speech Platform - Runtime Languages (2025-2026)

Runtime languages are powerful but operationally heavy. Use Speech Platform only for legacy on-prem needs — otherwise, move to Azure.

| Locale | Language | TTS Voices (Examples) | SR Model | |--------|----------|----------------------|-----------| | en-US | English (US) | Microsoft Zira, David | MSSpeech_SR_en-US_TELE.msi | | zh-CN | Chinese (Simplified) | Microsoft Hui, Kangkang | MSSpeech_SR_zh-CN_TELE.msi | | de-DE | German | Microsoft Hedda, Stefan | MSSpeech_SR_de-DE_TELE.msi | | fr-FR | French | Microsoft Hortense | MSSpeech_SR_fr-FR_TELE.msi | microsoft speech platform - runtime languages

using Microsoft.Speech.AudioFormat; using Microsoft.Speech.Recognition; Runtime languages are powerful but operationally heavy

Here is a breakdown of the essential information, structured as a technical brief, along with links to the official Microsoft documentation. Both SR and TTS functions work entirely offline

Both SR and TTS functions work entirely offline once the language packs are installed.

The "Runtime Languages" package is a critical component. Without it, the Speech Platform Runtime is an empty shell—it contains the engine to process audio but lacks the data files (acoustic models, lexicons, and voices) required to understand or speak specific languages.

You cannot just “set a language code.” You must install the corresponding for that language and locale.