Rime TTS

Arcana and Mist model families via Rime’s HTTP endpoint.

Rime TTS

RimeTTS targets the Rime HTTP endpoint. Both Arcana (high fidelity) and Mist (low latency) model families are supported. The provider requests audio/pcm and yields raw PCM_S16LE chunks at the configured sample rate.

Install

pip install "getpatter[rime]"

Usage

# Namespaced import
from getpatter.tts import rime

tts = rime.TTS()                                          # reads RIME_API_KEY
tts = rime.TTS(model="arcana", speaker="astra")

# Flat alias (equivalent)
from getpatter import RimeTTS

tts = RimeTTS(model="arcana", speaker="astra")

In an agent:

import asyncio
from getpatter import Patter, Twilio, DeepgramSTT, RimeTTS

phone = Patter(carrier=Twilio(), phone_number="[phone]")

agent = phone.agent(
    stt=DeepgramSTT(),                                    # DEEPGRAM_API_KEY from env
    tts=RimeTTS(model="arcana", speaker="astra"),         # RIME_API_KEY from env
    system_prompt="You are a helpful assistant.",
)

asyncio.run(phone.serve(agent))

Mist v2 — low-latency, streaming-friendly:

from getpatter import RimeTTS

fast = RimeTTS(model="mistv2", speaker="cove", speed_alpha=1.1, reduce_latency=True)

Low-level usage

from getpatter.providers.rime_tts import RimeTTS as _LowLevelTTS

tts = _LowLevelTTS(
    api_key="...",           # or RIME_API_KEY env var
    model="arcana",
    speaker="astra",
    lang="eng",
    sample_rate=16000,
)

async for chunk in tts.synthesize("Hello from the Patter pipeline."):
    ...

await tts.close()

Options (selection)

Option	Default	Notes
api_key	None	Reads from RIME_API_KEY when omitted.
model	"arcana"	"arcana" or any mist* model.
speaker	"astra" / "cove"	Default depends on model.
lang	"eng"	Rime language code.
sample_rate	16000	Hz.
temperature, top_p, max_tokens, repetition_penalty	—	Arcana only.
speed_alpha, reduce_latency, pause_between_brackets, phonemize_between_brackets	—	Mist only.