livekit-plugins-gnani

LiveKit Agents plugin for Gnani — high-accuracy Speech-to-Text (Prisma) and low-latency Text-to-Speech (Timbre) for Indian languages.

Gnani is a production-ready speech AI platform supporting 10+ Indian languages, real-time streaming, and multilingual transcription.

This integration is maintained by Gnani.ai.

Installation

pip install livekit-plugins-gnani

Or with uv:

uv add livekit-plugins-gnani

This will also install the websockets and livekit-agents packages as dependencies.

Install with the LiveKit Agents Gnani extra:

uv add "livekit-agents[gnani]"

Prerequisites

You need a Gnani API key. Gnani APIs have this.

Set your credentials as environment variables:

export GNANI_API_KEY="your-api-key"

Or pass the key in the constructor:

stt = STT(api_key="your-api-key", language="hi-IN")
tts = TTS(api_key="your-api-key")

Environment variables

Quick Start — AgentSession snippet

The snippet below shows core AgentSession wiring with Gnani WebSocket STT/TTS.

from livekit.agents import AgentSession, room_io
from livekit.plugins import gnani, groq, silero

session = AgentSession(
    stt=gnani.STT(
        language="en-IN",
        use_streaming=True,
    ),
    llm=groq.LLM(model="llama-3.1-8b-instant"),
    tts=gnani.TTS(
        voice="Nalini",
        model="timbre-v2.5",
        language="en-IN",
        synthesize_method="websocket",
    ),
    vad=silero.VAD.load(),
)

await session.start(
    agent=MyAgent(),
    room=ctx.room,
    room_options=room_io.RoomOptions(
        audio_input=room_io.AudioInputOptions(sample_rate=16000),
        audio_output=room_io.AudioOutputOptions(sample_rate=16000),
    ),
)

Set use_streaming=False (STT) or swap synthesize_method (TTS) for REST variants — see below. WebSocket STT + TTS is the default for lowest latency.

Service Construction

Speech-to-Text (REST)

from livekit.plugins.gnani import STT

stt = STT(
    language="hi-IN",
    use_streaming=False,
)

REST mode requires a VAD in the pipeline. LiveKit wraps the STT with stt.StreamAdapter automatically when use_streaming=False.

Speech-to-Text (Streaming WebSocket)

from livekit.plugins.gnani import STT

stt = STT(
    language="hi-IN",
    use_streaming=True,
    sample_rate=16000,
)

Text-to-Speech (REST)

from livekit.plugins.gnani import TTS

tts = TTS(
    voice="Pranav",
    model="timbre-v2.0",
    synthesize_method="rest",
)

Text-to-Speech (SSE Streaming)

from livekit.plugins.gnani import TTS

tts = TTS(
    voice="Pranav",
    synthesize_method="sse",
)

Text-to-Speech (WebSocket Streaming)

from livekit.plugins.gnani import TTS

tts = TTS(
    voice="Nalini",
    model="timbre-v2.5",
    language="hi-IN",
    synthesize_method="websocket",
)

The stream() method always uses WebSocket regardless of synthesize_method.

Services

STT

Streaming PCM Specification

All streaming audio must be sent as raw PCM binary frames — no container format (WAV, MP3) mid-stream.

Frames must be sent at real-time cadence. See STT Realtime — PCM Specification for full details.

TTS

Full Constructor Reference

STT — All parameters

from livekit.plugins.gnani import STT

stt = STT(
    language="en-IN",              # Default: "en-IN"
    sample_rate=16000,             # Default: 16000 (also: 8000, 44100, 48000)
    format="verbatim",             # Default: "verbatim" (also: "transcribe" for ITN)
    itn_native_numerals=False,     # Default: False
    use_streaming=True,            # Default: True (WebSocket); False = REST + VAD
    api_key=None,                  # Default: None (reads GNANI_API_KEY env var)
    base_url="https://api.vachana.ai",
)

TTS — All parameters

from livekit.plugins.gnani import TTS

tts = TTS(
    voice="Pranav",                # Default: "Pranav" (timbre-v2.0: Kaveri, Shubhra, Deepak)
    model="timbre-v2.0",           # Default: "timbre-v2.0" (also: "timbre-v2.5" with 42 voices)
    language=None,                 # timbre-v2.5 only — e.g. "hi-IN", "en-IN"
    sample_rate=16000,             # Default: 16000 (also: 8000, 22050, 44100)
    encoding="linear_pcm",         # Default: "linear_pcm" (also: "oggopus")
    container="wav",               # Default: "wav" (also: "raw", "mp3", "ogg")
    num_channels=1,                # Default: 1
    bitrate=None,                  # Default: None (also: "96k", "128k", "192k")
    synthesize_method="rest",      # Default: "rest" (also: "sse", "websocket")
    api_key=None,                  # Default: None (reads GNANI_API_KEY env var)
    base_url="https://api.vachana.ai",
)

Supported Languages

STT Languages (Prisma)

STT uses BCP-47 locale codes (e.g. hi-IN, bn-IN).

For the full list of supported languages, see:

• STT REST — Supported Languages
• STT Realtime — Supported Languages

TTS Languages (Timbre)

The optional language parameter is supported for timbre-v2.5 only. For the full list, see TTS — Supported Languages.

Migration: The former model name vachana-voice-v3 has been renamed to timbre-v2.0. Update any model="vachana-voice-v3" calls to model="timbre-v2.0" (or omit model to use the default).

Available Voices

See the official voice list for the latest supported voices.

timbre-v2.0 (4 voices)

timbre-v2.5 (42 voices)

The expanded catalog includes voices across Hindi, English, Tamil, Telugu, Kannada, Malayalam, Marathi, Bengali, Gujarati, Punjabi, and Hinglish. Use model="timbre-v2.5" with an optional language parameter (e.g. language="hi-IN").

from livekit.plugins.gnani import TTS

tts = TTS(model="timbre-v2.5", voice="Nalini", language="hi-IN")

Architecture

livekit-plugins-gnani    ← This package (LiveKit Agents adapter)
  ├── STT: REST + WebSocket
  └── TTS: REST + SSE + WebSocket

This plugin directly implements the Gnani REST and WebSocket APIs using aiohttp (for REST STT/TTS) and websockets (for streaming STT/TTS), adapting them into LiveKit's stt.STT and tts.TTS base classes. It uses the Prisma model for speech-to-text and the Timbre model for text-to-speech. No external SDK is required — all connection logic, authentication, and audio format handling is self-contained. Authentication uses a single api_key passed via the X-API-Key-ID header.

Documentation

• Gnani API Docs
• LiveKit Agents Docs
• Gnani STT Plugin Guide
• Gnani TTS Plugin Guide
• STT REST API
• STT Realtime WebSocket
• TTS REST API
• TTS Streaming (SSE)
• TTS Realtime WebSocket

LiveKit Compatibility

Tested with LiveKit Agents v1.6.x.

License

Apache 2.0 — see LICENSE.