VoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice design, video dubbing, dictation, transcription & audiobook creation in 646 languages.
Python aiaudiobookcudadubbingelevenlabs-alternativehuggingfacelocal-firstmlxomnivoice-studiospeech-to-texttext-to-speechtranscriptiontranslatettsvoice-aivoice-cloningvoice-generationvoicestudioworkflow
"""Dictation judges — the streaming-ASR (real-time capture) WebSocket contract.
Model-free: verifies the endpoint is registered and accepts a loopback handshake.
Driving real partial/final transcription needs an ASR model and is left to an
enable-on-demand live test.
"""
from __future__ import annotations
from ..spec import JudgeResult
def ws_endpoint_registered(ws_routes: list, path: str) -> JudgeResult:
routes = list(ws_routes or [])
ok = path in routes
return JudgeResult(
name="ws_endpoint_registered",
passed=ok,
measured=path,
detail=f"WebSocket {path!r} registered" if ok else f"{path!r} not in WS routes {routes}",
)
def ws_handshake_ok(connected: bool) -> JudgeResult:
ok = bool(connected)
return JudgeResult(
name="ws_handshake_ok",
passed=ok,
measured=ok,
detail="dictation WS accepted a loopback handshake" if ok
else "dictation WS refused the loopback handshake",
)