localsoundsapi

About

Project URL Category Python Project Title Portable offline AI audio studio for Windows Project Description A portable Windows audio studio that runs text-to-speech, voice cloning, transcription, music and sound-effect generation locally from one folder. It has a browser UI and a local REST API, and works offline once the models are downloaded. Highlights One-click setup: install.bat fetches a portable Python environment and tools, verifies SHA-256 hashes and extracts them. No system Python, conda or admin rights needed. Several engines in one place: XTTS v2 and Fish Speech (voice cloning), Kokoro 82M, Whisper (transcription and per-chunk verification), Stable Audio Open and ACE-Step (music and sound effects). Every model can be placed on CPU or any GPU separately, and a tkinter launcher runs multiple instances on different ports. Post-processing on every engine: de-reverb, de-essing, loudness normalization, silence trimming, with optional Whisper checks and automatic retries. Extras: a llama.cpp chatbot for writing scripts and lyrics, and a tool that turns audio plus transcript into a subtitled video. MIT licensed; Flask backend, so the REST API is easy to call from scripts. Example Code bat rem After cloning or downloading the repo ZIP: install.bat rem Later starts: launcher.bat Screenshots or Demo Videos No response Disclosure: I'm the author of this project (built with AI tools). Self-recommended.

Discovery history

1. HelloGitHub·用户投稿

   Oct 10, 2026 · 1 observations