livekit-plugins-speechmatics

What it is and what it does

This package is a plugin that bridges Speechmatics' speech-to-text service into the LiveKit Agents framework, enabling voice transcription in real-time agent applications. It wraps the Speechmatics API client and integrates it with LiveKit's agent session lifecycle, handling audio streaming, transcription callbacks, and speaker identification.

The plugin supports multiple turn-detection strategies: external (driven by VAD or manual finalization), adaptive (Speechmatics-controlled), smart turn (ML-based), and fixed silence duration. It can emit speaker identification metadata in configurable formats and accepts custom vocabulary entries for domain-specific terms. Configuration is done through the STT constructor, and the plugin requires a Speechmatics API key at runtime.

Use it for:

• Build a real-time conversational AI agent that transcribes user speech and identifies which participant is speaking in multi-party calls.
• Integrate Speechmatics transcription into a LiveKit-based meeting assistant that needs speaker attribution for accurate conversation logging.
• Create a voice-interactive application where turn-taking is controlled by Speechmatics' own endpointing rather than external VAD.
• Add domain-specific vocabulary (e.g., product names, technical terms) to transcription via the additional_vocab parameter.
• Deploy a LiveKit agent that uses Speechmatics for STT and delegates turn detection to the speech service itself.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Integrates Speechmatics speech-to-text into LiveKit Agents, enabling real-time transcription with optional speaker diarization and configurable turn detection modes.

Yes. The package is actively maintained (released 1 day ago), has no known vulnerabilities, low install friction, and Apache-2.0 licensing. Install it if you are building a LiveKit agent and need Speechmatics transcription with speaker diarization or flexible turn detection. Requires Python 3.10+ and a valid Speechmatics API key.

Install

livekit-plugins-speechmatics on PyPI

pip install livekit-plugins-speechmatics

uv add livekit-plugins-speechmatics

poetry add livekit-plugins-speechmatics

Installing livekit-plugins-speechmatics

Before you install

Low install friction; pure Python wheel with only two runtime dependencies (livekit-agents and speechmatics-voice). Package is actively maintained with a release 1 day old and no known vulnerabilities.

License in practice

Apache-2.0 permissive license allows commercial and private use with minimal restrictions; suitable for most production deployments.

Quickstart

pip install livekit-plugins-speechmatics

from livekit.agents import AgentSession
from livekit.plugins import speechmatics

agent = AgentSession(
    stt=speechmatics.STT(
        turn_detection_mode=speechmatics.TurnDetectionMode.ADAPTIVE,
        speaker_active_format="[Speaker {speaker_id}] {text}",
    ),
    turn_detection="stt",
)

Requires Speechmatics API key set as SPEECHMATICS_API_KEY environment variable or in .env.local file; Python 3.10 or later required.

Verify before relying

• Exact latency or throughput characteristics for real-time transcription under typical network conditions.
• Behavior and fallback strategy if Speechmatics API becomes unavailable mid-session.
• Compatibility with specific versions of livekit-agents and speechmatics-voice beyond the declared runtime dependencies.

Package facts

Evidence: livekit_plugins_speechmatics-1.6.10-py3-none-any.whl

Keywords: ai, audio, livekit, realtime, speechmatics, video, voice

Tags

More Artificial Intelligence packages

Further reading