Open source. Open possibilities.

A Python-based LLM routing system that categorizes queries by complexity to optimize API costs using a three-tier architecture across Gemini and Claude models.

Arpege is a top-down bullet hell RPG in a medieval fantasy setting, currently in early development focusing on its engine foundations.

GPT-SoVITS is an open-source few-shot voice conversion and TTS system supporting zero-shot (5s) and few-shot (1min) voice cloning across Chinese, Japanese, Korean, Cantonese, and English. Includes WebUI tools for audio separation, ASR, and dataset preparation.

Language not supported by the system.

Mosh is a native, cross-platform Digital Audio Workstation (DAW) that integrates AI directly into the signal path, featuring a generative timbre-transform layer and an in-app agent.

A sound effects plugin for DeepSeek Harness providing ambient work music and notification chimes.

LX Music 桌面版是一款基于 Electron 和 Vue 的跨平台音乐播放软件，支持 Windows、macOS、Linux，提供 Scheme URL 调用、开放 API、多端数据同步与自定义音源等功能。

Audiobookshelf is an open-source, self-hosted server for managing and streaming audiobooks and podcasts.

PeerLink X is a cross-platform Flutter messenger using decentralized networking. It features WebRTC for peer sessions and TURN for calls, relay-based messaging, and app-level cryptography. Supports self-hosting, end-to-end encrypted group chats, and privacy-focused design with internal contacts.

PWA Karaoke is a browser-based karaoke application that turns any device into a full karaoke system. Use phones or tablets as controllers to search and queue songs, while TVs or projectors serve as displays with synchronized playback.

Navidrome is an open-source, web-based music collection server and streamer that allows users to host and access their own music library from any browser or mobile device.

fizzle converts WAV/SFZ files into Casio FZ sampler disk images. Supports CLI and browser editor for creating, editing, and managing floppy disk images for FZ-1/10M/20M samplers.

A local-first video editing pipeline. It handles everything from transcription and silence detection to LLM-generated cut suggestions and final rendering via human approval, prioritizing privacy by keeping media on the PC.

Undertone is a dependency-free procedural synth engine for games and music, built on the Web Audio API to create one-shot sound effects, loops, and full songs using a pattern-based approach.

A rule-based classical piano music generator that creates complete pieces in the browser from a single seed.

Masters Codex is a digital tool suite for TTRPG Game Masters and Players, focused on the D&D 5th Edition system, emphasizing worldbuilding and mobile-first character sheets.

NVIDIA NeMo Speech is an Apache-2.0 PyTorch framework for building, customizing and deploying speech AI models, covering automatic speech recognition, text-to-speech and speech LLMs, with pre-trained checkpoints and Docker/uv install paths.

Valis is a virtual analog synthesis system that lets users build circuits using RDF/Turtle syntax, with UI views for controls, circuit diagrams, and code. It supports LLM-driven design via MCP and runs as a standalone app or DAW plugin.

Self-hosted web toolkit with client-side processing for images, PDFs, text, math, and data tools. Lightweight Docker image, no tracking, open-source under MIT.

A comprehensive online toolkit designed for developers and IT professionals, featuring 114+ tools. Offers practical functions including Traditional-Simplified conversion, QR code generation, image processing, PDF utilities, and P2P file transfer. Supports SEO optimization and direct AI Agent invocation, with a default Traditional Chinese interface optimized for users in Taiwan and Hong Kong.