LiveKit Agents Plugins Rime

1.5.7 verified Fri May 01 auth: no python

A LiveKit Agents plugin for Rime, providing text-to-speech TTS capabilities within the LiveKit realtime AI agent framework. Current version 1.5.7, requires Python >=3.10. Active development with frequent releases.

pip install livekit-plugins-rime

Common errors

error ModuleNotFoundError: No module named 'livekit' ↓

cause Missing metadata or incorrect installation of livekit-agents base package.

fix

Install livekit-agents: pip install livekit-agents

error Object of type 'coroutine' has no attribute 'data' ↓

cause Forgetting to await or use async context manager when consuming TTS stream.

fix

Use async with tts.synthesize(text) as stream: and then iterate with async for chunk in stream:

error RuntimeError: Rime API key not provided ↓

cause Neither `RIME_API_KEY` environment variable nor `api_key` arg was set.

fix

Set environment variable or pass api_key parameter explicitly.

Warnings

breaking Version 1.0+ changed the TTS interface to async generator streams. Old synchronous `synthesize()` method removed. ↓

fix Use `async with tts.synthesize(text) as stream` and iterate over chunks.

deprecated The `voice` parameter in RimeTTS is deprecated in favor of `voice_id`. Passing `voice` will raise a deprecation warning. ↓

fix Replace `voice="..."` with `voice_id="..."`.

gotcha The Rime API key must be set via environment variable `RIME_API_KEY` or passed in constructor. If neither is found, the plugin fails silently. ↓

fix Always either set `RIME_API_KEY` env var or provide `api_key` argument.

Imports

RimeTTS
wrong
```
from livekit_plugins_rime import RimeTTS
```
correct
```
from livekit.plugins.rime import RimeTTS
```
Old import paths with underscores are incorrect; the official namespace uses dots.

TTS

wrong

from livekit.plugins.rime import TTS

correct

from livekit.agents.tts import TTS

TTS base class is from agents, not the plugin.

Quickstart

Basic usage of RimeTTS to synthesize speech, assuming API key is set in environment.

import asyncio
from livekit.plugins.rime import RimeTTS

async def main():
    tts = RimeTTS(
        api_key=os.environ.get("RIME_API_KEY", ""),
        voice_id="test-voice"
    )
    async with tts.synthesize("Hello, world!") as stream:
        async for chunk in stream:
            print(chunk.data)

asyncio.run(main())