Lalk

Lalk

Real-time voice agents, built for natural conversation.

A Python SDK, server, and desktop workspace for real-time voice agents.

Python PyPI Platform

English | 简体中文

Overview

Lalk brings the complete real-time voice pipeline into one project: microphone capture, voice activity detection, turn detection, speech recognition, Agent execution, speech synthesis, interruption handling, observability, a local server, and a Tauri desktop application.

The Python SDK is modular, so each audio, ASR, VAD, Agent, TTS, and turn-detection component can be replaced independently. The desktop application packages the Python server as a local sidecar and keeps runtime traffic on localhost.

Core Capabilities

  • Full-duplex voice sessions with natural user interruption.
  • Native macOS VoiceProcessingIO audio with echo cancellation.
  • Silero VAD, adaptive input gating, and Smart Turn semantic turn detection.
  • Local ASR and interchangeable cloud speech recognition implementations.
  • Bumblehive-powered Agent runtime with streaming events and tool execution.
  • Streaming TTS with word-level playback marks.
  • Conversation inactivity policies and proactive follow-up support.

Platform Support

Lalk currently targets Apple Silicon Macs (M-series) running macOS 14 or later.

Requirements

  • Apple Silicon Mac
  • macOS 14+
  • Python 3.11+
  • Node.js 22+
  • pnpm 10.33.0
  • Rust stable toolchain
  • Xcode Command Line Tools

Quick Start

Install the Python SDK from PyPI:

python -m pip install lalk

For repository development, create and activate a Python environment:

conda create -n lalk_env python=3.11 -y

Install all project dependencies and verify the environment:

pnpm run setup

The setup command prepares the Python and Node.js dependencies required for development and desktop packaging.

If downloading from the default PyPI index is slow, use the Tsinghua PyPI mirror for this setup command only:

PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple pnpm run setup

Development

Start the local server on 127.0.0.1:17841:

pnpm dev:server

In another terminal, start the WebUI:

pnpm dev:web

Start the desktop application in development mode:

pnpm dev:desktop

Build the Desktop App

Create the Apple Silicon macOS application and DMG installer:

pnpm build:desktop

The installer is written to:

desktop/src-tauri/target/aarch64-apple-darwin/release/bundle/dmg/

The build pipeline automatically:

  1. Compiles the native VoiceProcessingIO library.
  2. Packages the Python server as the lalk-server sidecar.
  3. Builds the React WebUI.
  4. Bundles and signs the Tauri application locally.
  5. Creates the Apple Silicon DMG.

Project links

Data verified by PyPI on Sep 3, 2026
Data provided by the project maintainers, verified at the time the release was uploaded to PyPI.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Latest release

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for wxhcore from gravatar.com wxhcore

Credits

Author: wxhcore

GitHub Statistics

Data verified by PyPI on Sep 3, 2026
The GitHub source repository was provided by the project maintainers and verified by PyPI at the time of upload. Stars, forks, and open issues/PRs are derived from that repository and have not been independently verified.

License expression

Apache-2.0
View SPDX License List

Requires

Python >=3.11

Provides Extra

build server test

Tags

ai audio realtime speech voice-agent

Classifiers

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

lalk-0.1.0.tar.gz (9.6 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

lalk-0.1.0-py3-none-any.whl (9.6 MB view details)

Uploaded Python 3

File details

Details for the file lalk-0.1.0.tar.gz.

File metadata

  • Download URL: lalk-0.1.0.tar.gz
  • Upload date:
  • Size: 9.6 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for lalk-0.1.0.tar.gz
Algorithm Hash digest
SHA256 5852a9c90f6357a340cf8422c5705e588e5e4ef28b059edcc96821172b95854b
MD5 2b338d41727bfbc2128a6a29fa8b174f
BLAKE2b-256 c4d301da55612f6ec1de3cbd49cb069026e204d5b8431ace906348960ab4d878

See more details on using hashes here.

Provenance

The following attestation bundles were made for lalk-0.1.0.tar.gz:

Publisher: sdk-release.yml on WithLing/lalk

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file lalk-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: lalk-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 9.6 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for lalk-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 5a2e3dc6ef9c4dde1c4e94468eef4a939c3ba82e9d0601061635c207ad151c24
MD5 cfb33ff7a42255648460b1f2eb9ac90d
BLAKE2b-256 826ca2225551d921b84c2225206e3dc59a3e7c8d0367f2d495b3d92e221d0b90

See more details on using hashes here.

Provenance

The following attestation bundles were made for lalk-0.1.0-py3-none-any.whl:

Publisher: sdk-release.yml on WithLing/lalk

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page