LocalSoundsAPI

LocalSoundsAPI is an all-in-one AI audio studio that runs entirely from one folder, with no installation and no internet needed after setup. It includes XTTS v2 and Fish Speech for voice cloning, Kokoro for fast English text-to-speech, Stable Audio Open and ACE-Step for music and sound effects, Whisper for transcription and subtitles, and a built-in local LLM chatbot for writing prompts, scripts and lyrics. Every output goes through de-reverb, de-essing, loudness normalization and silence trimming. Each model can be placed on the CPU or any GPU, several instances can run at once, and audio plus transcripts can be turned into subtitled videos. Use it from the browser interface or call the local REST API from your own scripts. MIT licensed.