Agent-readable docs index: /llms.txt. Download /docs.zip to grep all markdown files locally.

LiveKit Agents

Vakyam is an official LiveKit Agents TTS plugin (livekit-plugins-vakyam). Install it next to livekit-agents; it is versioned with the LiveKit Agents monorepo, independently of the core vakyamai SDK.
stream() uses the realtime WebSocket API and sentence-tokenizes LLM text so each utterance is one complete sentence. synthesize() uses HTTP streaming (POST /v1/tts/stream) and returns PCM audio.
The vakyamai[livekit] extra and from vakyamai.livekit import VakyamTTS remain available for now, but they will be deprecated. Switch to the official LiveKit plugin — the constructor arguments are the same; only the package and import path change. See Migrate from the SDK extra.

Install

pip install livekit-plugins-vakyam
Or with the LiveKit Agents extra:
uv add "livekit-agents[vakyam]"
Source and tests live in the livekit-plugins-vakyam package in the LiveKit Agents repository.
Set your API key:
export VAKYAM_API_KEY="vak_live_..."

Create the TTS plugin

from livekit.plugins import vakyam

tts = vakyam.TTS(
    model="raaga-v1",
    voice="Archana",
    language="ta-IN",
    sample_rate=24000,
    speed=1.0,
)
The plugin reads VAKYAM_API_KEY from the environment by default; pass api_key="vak_live_..." to set it explicitly.

Parameters

path api_keystring
Vakyam API key. Defaults to the VAKYAM_API_KEY environment variable.
path modelstringdefault: raaga-v1
TTS model identifier.
path voicestringdefault: Archana
Voice selector — a preset name (discover valid values with GET /v1/voices) or a custom voice ID beginning with vc_. See Voices & languages.
path languagestringdefault: ta-IN
BCP 47 language code, such as ta-IN, hi-IN, or en-IN.
path sample_rateintegerdefault: 24000
Output PCM sample rate in Hz: 8000, 16000, 24000, or 48000.
path speednumberdefault: 1.0
Speech rate multiplier, 0.5–2.0.
path base_urlstringdefault: https://api.vakyam.ai
API base URL. Defaults to production (https://api.vakyam.ai).
path allow_insecure_base_urlbooleandefault: false
Allow non-localhost http:// base URLs. Defaults to false.
path tokenizerSentenceTokenizer
Optional custom sentence tokenizer. Defaults to LiveKit's built-in sentence tokenizer.
Call update_options(...) to change model, voice, language, sample_rate, or speed for streams created after the call. Call prewarm() to open a pooled WebSocket before the first turn.

Use it in an AgentSession

Combine vakyam.TTS with your chosen STT, LLM, and VAD, the same way you would wire any other LiveKit TTS plugin:
from livekit.agents import AgentSession
from livekit.plugins import vakyam

session = AgentSession(
    stt=your_stt,
    llm=your_llm,
    tts=vakyam.TTS(
        model="raaga-v1",
        voice="Archana",
        language="ta-IN",
        sample_rate=24000,
    ),
    vad=your_vad,
)
vakyam.TTS supports streaming synthesis and interruption. WebSocket connections are pooled and reused between sequential agent turns. Each active synthesis stream has exclusive ownership of its connection, so an overlapping stream uses a separate connection. On barge-in, the plugin sends a cancel, drains through Vakyam's cancellation acknowledgement, and returns the healthy connection to the pool. See Interrupting playback.

Migrate from the SDK extra

If you still install vakyamai[livekit], switch the package and import. The constructor arguments are unchanged. Keep vakyamai installed if you use the core SDK elsewhere.
pip install livekit-plugins-vakyam
# Before (will be deprecated)
from vakyamai.livekit import VakyamTTS

tts = VakyamTTS(
    model="raaga-v1",
    voice="Archana",
    language="ta-IN",
    sample_rate=24000,
)

# After
from livekit.plugins import vakyam

tts = vakyam.TTS(
    model="raaga-v1",
    voice="Archana",
    language="ta-IN",
    sample_rate=24000,
)
Vakyam AI
Powered byHolocron