Rime

The RimeTTSService provides high-fidelity, low-latency text-to-speech using Rime's WebSocket API. It supports advanced features like inline speed control and specialized spelling modes.

Installation

pip install "piopiy-ai[rime]"

Prerequisites

• A Rime API key (Get yours here).
• Set your API key in your environment:

  export RIME_API_KEY="your_api_key_here"

Configuration

RimeTTSService Parameters

Parameter	Type	Default	Description
api_key	str	Required	Your Rime API key.
voice_id	str	Required	ID of the voice to use.
model	str	"mistv2"	Model ID for synthesis.
params	InputParams	InputParams()	Synthesis settings.

InputParams

Parameter	Type	Default	Description
language	Language	Language.EN	Language for synthesis.
speed_alpha	float	1.0	Speed multiplier.
reduce_latency	bool	False	Optimize for speed over quality.

Usage

Basic Setup

import os
from piopiy.services.rime.tts import RimeTTSService

tts = RimeTTSService(
    api_key=os.getenv("RIME_API_KEY"),
    voice_id="marcia"
)

Notes

• Word Timestamps: Rime provides precise word-level timing information, which is used for lip-sync and progress tracking.
• Special Functions: Supports SPELL(text) for spelling out words and PAUSE_TAG(seconds) for inserting pauses.
• Alternative Endpoints: For models like Arcana, use RimeNonJsonTTSService.