Model
Load TTS model
POST
Load or switch the TTS model.
A model must be loaded before any generation endpoint will work.
If the requested mode/quality is already loaded, returns immediately
with status: "already_loaded". Otherwise begins async loading —
poll GET /api/v1/model/load/progress until status: "complete".
Switching generation mode (clone / design / tts) requires a different model variant and triggers a reload.
sync?SyncWait for loading to complete
Default
falseapplication/json- body
Default
{}mode?|Generation mode to load (clone, design, tts). Uses current if omitted.
quality?|Quality preset to load. Uses current if omitted.
Successful Response
application/json- response
status*StatusLoad status (loading, complete, already_loaded, error)
mode?|Generation mode being loaded
quality?|Quality preset being loaded
message?|Error message if status is 'error'
curl -X POST 'http://127.0.0.1:8100/api/v1/model/load' \ -H 'Content-Type: application/json' \ -d '{}'Example Response
{ "status": "loading", "mode": "clone", "quality": "higher_quality", "message": "string"}