Features

Everything rAIdio.bot does.

One application. Every AI music feature you need. No cloud, no subscription.

Six-channel mixer with per-channel meters and a master bus
MixSix-channel mixer, per-channel FX, master bus with EQ and limiter.
Voice tab with modes dropdown
VoiceText-to-speech, voice design, cloning, conversion, and training.
Play tab with 3D waveform and Sing Along
Play3D waveform visualization, karaoke with word-level timing, video support.
Properties panel with C2PA Signed and Valid status
ProvenanceC2PA Content Credentials on every generated file.

Generate songs from text prompts

Describe what you want in plain English. rAIdio.bot generates a full, original song, vocals, instruments, structure, in seconds. Powered by ACE-Step, a diffusion model trained exclusively on licensed and public domain audio. Every song comes with a seed you can share or save for reproducibility.

Prompt Enhance

One click on Prompt Enhance and a local model rewrites your prompt into something the generator understands better. It runs on your own machine, so nothing leaves it.

Extend & Repaint

Grow a track past its original length when it ends too soon, or select a section and regenerate only that part. The rest of the track stays as it was.

Remix

Right-click any generated file to reload its exact settings: prompt, seed, steps, CFG, and the rest. Change what you want, re-render, compare.

Clone your voice from 30 seconds of audio

Record 30 seconds of your voice. rAIdio.bot clones it. Use the clone for TTS, or convert an existing recording into that voice. Works with your own voice, voice actors who've given consent, or any recording you have the rights to use. Every clone requires explicit consent attestation, built into the workflow, embedded in the output's provenance signature.

Text-to-speech in multiple voices

Type what you want said. Pick a voice. Hit generate. Natural-sounding speech in multiple built-in voices, perfect for narration, podcast intros, or adding spoken elements to your music.

Multi-speaker voice scripts

Write a script with a cast of distinct voices and render the whole narration in one pass. Set emotion per line, recast a voice, or re-render a single line without touching the rest. This is the same approach that keeps Doomscroll.fm on the air around the clock.

Emotion presets

Emotion and intensity presets for spoken and sung voice, so a line can be read calm, urgent, warm, or flat without rewriting it.

Train a custom voice model

Got 30 minutes of clean recordings? Train a full custom voice model. Training takes 1–3 hours on your GPU. The result is a personal voice model that produces the highest-quality, most consistent synthesis.

Separate songs into stems

Any song, split into vocals, drums, bass, and everything else. Right-click, extract stems. Meta Demucs running locally. Use the stems for remixes, karaoke tracks, sampling, or just to hear what's in your favorite songs.

Edit, mix, master on-device

Trim, cut, fade, normalize, apply 10 built-in effects (EQ, compressor, reverb, delay, bitcrusher, more). Real-time preview. Non-destructive. Your originals are never touched.

Six-channel mixer with master bus

Six channels. Drag stems in. Adjust volume, pan, apply per-channel effects. Master bus with EQ, compressor, saturation, stereo width. Export as WAV or MP3.

Karaoke mode with word-level timing

Feed it any song with vocals. rAIdio.bot separates the vocals, transcribes them with word-level timing, and plays the instrumental with synced lyrics highlighted word-by-word. Toggle the original vocals on and off. Works over video files too.

Generate soundtracks for your videos

Drag a video in. rAIdio.bot analyzes the energy and mood, consults an on-device AI music director, and generates a matching soundtrack. Entirely local.

Creative Director

Describe a mood in plain English, or hand it a video. An on-device AI plans the soundtrack: sections, energy, and instrumentation, then hands the plan to the generator.

Export audio as MIDI

Convert any audio clip to MIDI using Spotify's basic-pitch neural transcription. Highest accuracy on isolated melody and bass stems.

Voice conversion

A voice conversion model. Source audio in, processed audio out, timing preserved.

Chord detection

Feed any audio. rAIdio.bot detects the chord progression with timing, so you can jam along, transcribe, or remix in key.

Projects

Every track and session lives in its own project folder, and renders collect there on their own. No hunting through one flat output directory to find the take you liked.

All of this runs on your PC. None of it requires an internet connection.

A license key is required to run it: €19.99, one time.

Download rAIdio.bot Check system requirements