SAPIServer

Simple HTTP frontend API for Microsoft Speech API

Building

Requirements

• .NET SDK
• VS2022 lib pack (TODO: link)
• mingw-w64 toolchain built with win32 thread model (pthread won't work)

You'll also need to chattr +F (or mount the whole thing with ciopfs and rename the headers to lowercase, if not on ext4 or you don't want to tune2fs) the windows sdk header directories so the build works.

just should do the trick.

Running

Usage: SAPIServer.exe <port>

API Usage

List voices

GET /api/voices

{"voices":["LH Michael","LH Michelle","Microsoft Sam"]}

Synthesize text

POST /api/synthesize 
Content-Type: application/json

{"text":"Lorem ipsum doler sit amet...","voice":"Microsoft Sam"}

Returns synthesized TTS audio as a wave-form file.