refactor: restructure plugins by provider

Author: maxkahan

examples/llm_audio_conversation/main.py

@@ -28,9 +28,9 @@
 from openai import OpenAI
 
 from examples.utils import create_user, open_browser
-from getstream.plugins.stt.deepgram import Deepgram
-from getstream.plugins.tts.elevenlabs import ElevenLabs
-from getstream.plugins.vad.silero import Silero
+from getstream.plugins.deepgram import DeepgramSTT
+from getstream.plugins.elevenlabs import ElevenLabsTTS
+from getstream.plugins.silero import SileroVAD
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc import audio_track
@@ -73,11 +73,11 @@ async def main():
 
     # Initialize components
     audio = audio_track.AudioStreamTrack(framerate=16000)
-    vad = Silero()
-    stt = Deepgram()
+    vad = SileroVAD()
+    stt = DeepgramSTT()
 
     # Use "Arnold" voice - a default ElevenLabs voice available to all users
-    tts_instance = ElevenLabs(voice_id="VR6AewLTigWG4xSOukaG")  # ID of default voice
+    tts_instance = ElevenLabsTTS(voice_id="VR6AewLTigWG4xSOukaG")  # ID of default voice
     openai_client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
     tts_instance.set_output_track(audio)
 

examples/mcp/main.py

@@ -4,8 +4,8 @@
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.call import Call
-from getstream.plugins.stt.deepgram import Deepgram
-from getstream.plugins.tts.elevenlabs import ElevenLabs
+from getstream.plugins.deepgram import DeepgramSTT
+from getstream.plugins.elevenlabs import ElevenLabsTTS
 from getstream.video.rtc import audio_track
 from getstream.video.rtc.track_util import PcmData
 from examples.mcp.agent import chat_with_tools
@@ -20,8 +20,8 @@ async def run_bot(call: Call, bot_user_id: str):
     """Join the call as a bot, convert speech→text, call MCP tools
     when the transcript is final, then speak the answer back."""
 
-    stt = Deepgram()
-    tts = ElevenLabs()
+    stt = DeepgramSTT()
+    tts = ElevenLabsTTS()
     track = audio_track.AudioStreamTrack(framerate=16000)
     tts.set_output_track(track)
 

examples/openai_realtime_speech_to_speech/main.py

@@ -6,7 +6,7 @@
 from examples.utils import create_user, open_browser
 from getstream import Stream
 from getstream.models import StartClosedCaptionsResponse
-from getstream.plugins.sts.openai_realtime import OpenAIRealtime
+from getstream.plugins.openai import OpenAIRealtime
 
 
 logging.basicConfig(

examples/stt_deepgram_transcription/main.py

@@ -24,7 +24,7 @@
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc.track_util import PcmData
-from getstream.plugins.stt.deepgram import Deepgram
+from getstream.plugins.deepgram import DeepgramSTT
 from examples.utils import create_user, open_browser
 
 logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
@@ -70,7 +70,7 @@ async def main():
     print("\nPress Ctrl+C to stop the transcription bot.\n")
 
     # Initialize Deepgram STT (api_key comes from .env)
-    stt = Deepgram()
+    stt = DeepgramSTT()
 
     try:
         async with await rtc.join(call, bot_user_id) as connection:

examples/stt_moonshine_transcription/main.py

@@ -23,11 +23,11 @@
 import uuid
 from dotenv import load_dotenv
 
-from getstream.plugins.vad.silero.vad import Silero
+from getstream.plugins.silero import SileroVAD
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc.track_util import PcmData
-from getstream.plugins.stt.moonshine import Moonshine
+from getstream.plugins.moonshine import MoonshineSTT
 from examples.utils import create_user, open_browser
 
 
@@ -64,11 +64,11 @@ async def main() -> None:  # noqa: D401
     print("\n🤖 Starting transcription bot…")
     print("Speak in the browser and see transcripts below. Press Ctrl+C to stop.\n")
 
-    stt = Moonshine()
+    stt = MoonshineSTT()
 
     # Initialize Silero VAD for speech detection
     print("🔊 Initializing Silero VAD...")
-    vad = Silero()
+    vad = SileroVAD()
     print("✅ Audio processing pipeline ready: VAD → Moonshine STT")
 
     try:

examples/tts_cartesia/main.py

@@ -29,8 +29,7 @@
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc import audio_track
-from getstream.plugins.tts.cartesia import CartesiaTTS
-from getstream.video.rtc.pb.stream.video.sfu.models.models_pb2 import Participant
+from getstream.plugins.cartesia import CartesiaTTS
 
 
 logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
@@ -86,7 +85,7 @@ def _on_participant_joined(p):
 
             await someone_joined.wait()  # blocks here until a listener is present
 
-            # Now it’s safe to speak – nothing will be lost
+            # Now it's safe to speak – nothing will be lost
             await tts.send(greeting)
             logging.info("Sent greeting via TTS")
 

examples/tts_elevenlabs/main.py

@@ -29,7 +29,7 @@
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc import audio_track
-from getstream.plugins.tts.elevenlabs import ElevenLabs
+from getstream.plugins.elevenlabs import ElevenLabsTTS
 
 
 logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
@@ -60,7 +60,7 @@ async def main() -> None:
     open_browser(client.api_key, token, call_id)
 
     track = audio_track.AudioStreamTrack(framerate=16000)
-    tts = ElevenLabs()  # API key picked from ELEVENLABS_API_KEY
+    tts = ElevenLabsTTS()  # API key picked from ELEVENLABS_API_KEY
     tts.set_output_track(track)
 
     greeting = "Hello there! I'm an ElevenLabs T T S bot speaking inside this call. As this is a minimal example, I'll stop speaking now."

examples/vad_silero/main.py

@@ -28,7 +28,7 @@
 from getstream.stream import Stream
 from getstream.video import rtc
 from getstream.video.rtc.track_util import PcmData
-from getstream.plugins.vad.silero import Silero
+from getstream.plugins.silero import SileroVAD
 
 # ---------------------------------------------------------------------------
 # Logging setup – INFO level so we see joins / leaves, etc.
@@ -63,7 +63,7 @@ async def main() -> None:
 
     open_browser(client.api_key, token, call_id)
 
-    vad = Silero()
+    vad = SileroVAD()
 
     print("\n🤖 VAD bot starting – speak in the call and watch the console.\n")
 

getstream/plugins/tts/cartesia/README.md getstream/plugins/cartesia/README.mdgetstream/plugins/tts/cartesia/README.md renamed to getstream/plugins/cartesia/README.md

@@ -0,0 +1,3 @@
+from getstream.plugins.cartesia.tts import CartesiaTTS
+
+__all__ = ["CartesiaTTS"]