Lists (13)
Starred repositories
A fast native Rust Mermaid diagram renderer. No browser required. 500-1000x faster than mermaid-cli.
🌋LavaSR: Fast Speech restoration and enhancement
FineTune, a macOS menu bar app for per-app volume control, multi-device output, audio routing, and 10-band EQ. Free and open-source alternative to SoundSource.
An efficient, reliable parser for CommonMark, a standard dialect of Markdown
Streaming speech recognition running natively and in the browser. A pure Rust implementation of Mistral's Voxtral Mini 4B Realtime model using the Burn ML framework.
Open-source, customer support platform with fully customizable AI support agents for developers / startups shipping SaaS.
Powerful, lightweight web terminal. Batteries included. Powered by libghostty-vt, WebGPU, and text-shaper.
A community trust management system based on explicit vouches to participate.
Speech To Speech: an effort for an open-sourced and modular GPT4-o
Voice-to-text app for macOS to transcribe what you say to text almost instantly
Frontier CoreML audio models in your apps — text-to-speech, speech-to-text, voice activity detection, and speaker diarization. In Swift, powered by SOTA open source.
Pure C inference of Mistral Voxtral Realtime 4B speech to text model
Code's quiet companion. A minimalist theme for Neovim, written in Lua.
The most powerful local music generation model that outperforms most commercial alternatives, supporting Mac, AMD, Intel, and CUDA devices.
This library strips unwanted tracking and clutter from URLs, making them simpler, tidier, and easier to read.
A tool for creating and running Linux containers using lightweight virtual machines on a Mac. It is written in Swift, and optimized for Apple silicon.
A lightweight alternative to Clawdbot / OpenClaw that runs in containers for security. Connects to WhatsApp, has memory, scheduled jobs, and runs directly on Anthropic's Agents SDK
A professional-grade interface for Qwen3-TTS, designed to unlock the model's full potential with fine-grained control and intuitive workflows.
Fast, Sharp & Reliable Agentic Intelligence
A free, open source, and extensible speech-to-text application that works completely offline.
Qwen3-0.6B megakernel: 527 tok/s decode on RTX 3090 (3.8x faster than PyTorch)
Generalist and Lightweight Model for Named Entity Recognition (Extract any entity types from texts) @ NAACL 2024