LiveKit

1
from dotenv import load_dotenv
2
3
from livekit import agents
4
from livekit.agents import Agent, AgentServer, AgentSession
5
from livekit.plugins import deepgram, openai, speechify
6
7
load_dotenv()
8
9
10
class Assistant(Agent):
11
    def __init__(self) -> None:
12
        super().__init__(
13
            instructions=(
14
                "You are a helpful voice assistant speaking with a Speechify voice. "
15
                "Keep replies short, clear, and conversational."
16
            )
17
        )
18
19
20
server = AgentServer()
21
22
23
@server.rtc_session(agent_name="speechify-tts-demo")
24
async def entrypoint(ctx: agents.JobContext) -> None:
25
    session = AgentSession(
26
        stt=deepgram.STT(model="nova-3"),
27
        llm=openai.LLM(model="gpt-4o-mini"),
28
        tts=speechify.TTS(voice_id="dominic_32", model="simba-3.2"),
29
    )
30
31
    await session.start(room=ctx.room, agent=Assistant())
32
    await session.generate_reply(
33
        instructions="Greet the user and mention that your voice is powered by Speechify TTS."
34
    )
35
36
37
if __name__ == "__main__":
38
    agents.cli.run_app(server)

1
import asyncio
2
import wave
3
4
from livekit.plugins import speechify
5
6
7
async def main() -> None:
8
    tts = speechify.TTS(voice_id="dominic_32", model="simba-3.2")
9
    frames = []
10
    async for event in tts.synthesize("Hello from Speechify TTS on LiveKit."):
11
        frames.append(event.frame)
12
13
    with wave.open("speechify-tts-smoke.wav", "wb") as wf:
14
        wf.setnchannels(1)
15
        wf.setsampwidth(2)
16
        wf.setframerate(24000)
17
        for frame in frames:
18
            wf.writeframes(frame.data)
19
20
21
asyncio.run(main())