Inworld plugin for LiveKit Agents
Support for voice synthesis and speech-to-text with Inworld TTS and Inworld STT.
See Inworld TTS and Inworld STT for more information.
Installation
pip install livekit-plugins-inworld
Authentication
Set INWORLD_API_KEY in your .env file (get one here).
Usage
TTS
Use Inworld TTS within an AgentSession or as a standalone speech generator.
from livekit.plugins import inworld
tts = inworld.TTS()
Or with options:
from livekit.plugins import inworld
tts = inworld.TTS(
voice="Hades", # voice ID (default or custom cloned voice)
model="inworld-tts-1", # or "inworld-tts-1-max"
encoding="OGG_OPUS", # LINEAR16, MP3, OGG_OPUS, ALAW, MULAW, FLAC
sample_rate=48000, # 8000-48000 Hz
bit_rate=64000, # bits per second (for compressed formats)
speaking_rate=1.0, # 0.5-1.5
temperature=1.1, # 0-2
timestamp_type="WORD", # WORD, CHARACTER, or TIMESTAMP_TYPE_UNSPECIFIED
text_normalization="OFF", # ON, OFF, or APPLY_TEXT_NORMALIZATION_UNSPECIFIED
)
TTS Streaming
Inworld TTS supports WebSocket streaming for lower latency real-time synthesis. Use the
stream() method for streaming text as it's generated:
from livekit.plugins import inworld
tts = inworld.TTS(
voice="Hades",
model="inworld-tts-1",
buffer_char_threshold=100, # chars before triggering synthesis (default: 100)
max_buffer_delay_ms=3000, # max buffer time in ms (default: 3000)
)
# Create a stream for real-time synthesis
stream = tts.stream()
# Push text incrementally
stream.push_text("Hello, ")
stream.push_text("how are you today?")
stream.flush() # Flush any remaining buffered text
stream.end_input() # Signal end of input
# Consume audio as it's generated
async for audio in stream:
# Process audio frames
pass
STT
Use Inworld STT for streaming speech-to-text. Multiple models are supported.
from livekit.plugins import inworld
session = AgentSession(
stt=inworld.STT()
# ... llm, tts, etc.
)
With a specific model and voice profile detection:
from livekit.plugins import inworld
session = AgentSession(
stt=inworld.STT(
model="inworld/inworld-stt-1",
enable_voice_profile=True,
)
# ... llm, tts, etc.
)
Example
A full voice agent using Inworld for both STT and TTS:
"""Inworld STT + TTS voice agent example.
Demonstrates using Inworld for both speech-to-text and text-to-speech
in a LiveKit voice agent. Save this as ``inworld_agent.py`` and run:
uv run inworld_agent.py console # local console mode
uv run inworld_agent.py dev # LiveKit Cloud (requires LIVEKIT_URL,
# LIVEKIT_API_KEY, LIVEKIT_API_SECRET)
Then connect via https://agents-playground.livekit.io
"""
import logging
from dotenv import load_dotenv
from livekit.agents import (
Agent,
AgentServer,
AgentSession,
JobContext,
cli,
inference,
metrics,
room_io,
)
from livekit.agents.inference import TurnDetector
from livekit.plugins import inworld
logger = logging.getLogger("inworld-agent")
load_dotenv()
class InworldAgent(Agent):
def __init__(self) -> None:
super().__init__(
instructions=(
"Your name is Nova. You interact with users via voice. "
"Keep your responses concise and to the point. "
"Do not use emojis, asterisks, markdown, or other special characters. "
"You are helpful, curious, and friendly."
),
)
async def on_enter(self):
self.session.generate_reply()
server = AgentServer()
@server.rtc_session()
async def entrypoint(ctx: JobContext):
ctx.log_context_fields = {"room": ctx.room.name}
session = AgentSession(
stt=inworld.STT(model="inworld/inworld-stt-1"),
llm="openai/gpt-4.1-mini",
tts=inworld.TTS(voice="Clive"),
turn_detection=TurnDetector(),
vad=inference.VAD(),
)
usage_collector = metrics.UsageCollector()
@session.on("metrics_collected")
def _on_metrics(ev):
metrics.log_metrics(ev.metrics)
usage_collector.collect(ev.metrics)
async def log_usage():
logger.info(f"Usage: {usage_collector.get_summary()}")
ctx.add_shutdown_callback(log_usage)
await session.start(
agent=InworldAgent(),
room=ctx.room,
room_options=room_io.RoomOptions(),
)
if __name__ == "__main__":
cli.run_app(server)
Combined TTS + STT
from livekit.plugins import inworld
session = AgentSession(
tts=inworld.TTS(voice="Hades"),
stt=inworld.STT(),
# ... llm, etc.
)
Release files for livekit-plugins-inworld 1.8.3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| livekit_plugins_inworld-1.8.3.tar.gz | 22.0 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| livekit_plugins_inworld-1.8.3-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 46.0 kB
Release files / livekit_plugins_inworld-1.8.3.tar.gz
| Download URL | livekit_plugins_inworld-1.8.3.tar.gz |
|---|---|
| Size | 22.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
0638b9b9e00ec6d2f09fe923f33ea8f8810dc80125d1a5bbc2dc1f579060356e
|
|
BLAKE2b-256 checksum How to use checksums |
9759d31f639a0cdb61cd4a3a51e732827c3c75095decc3a0c2c9ee858581b262
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 23, 2026.
Transparency logRelease files / livekit_plugins_inworld-1.8.3-py3-none-any.whl
| Download URL | livekit_plugins_inworld-1.8.3-py3-none-any.whl |
|---|---|
| Size | 23.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
7f34072ccbcea111b42e119e38337224f38ff762a41bc7778e58cd462221a977
|
|
BLAKE2b-256 checksum How to use checksums |
4283283db7c246107d29fbd7b721f64efefde4ef658029869231e65c6cc3826c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 23, 2026.
Transparency log
