Offline text-to-speech pipeline

🎙️ Vox

What it is

Vox turns text into speech entirely on local hardware. It takes whatever the system needs to say — output, a nudge, an ambient note — and renders it as audio using a neural voice model running on the machine itself. No API calls, no per-request fees, and no audio ever leaves the box. One command, say "…", and a voice comes out.

Most "natural AI voice" setups reach for a cloud TTS API, which buys quality at the cost of latency, a per-word bill, an internet dependency, and a third party hearing everything. Vox takes the opposite bet: own the whole pipeline, keep it offline, and make it fast enough to feel conversational.

One command, one voice, no cloud. The voice lives on the machine — nothing to call, nothing to leak.

The philosophy

Local-first Zero network dependency for synthesis — it works on a plane, in a blackout, behind any firewall.
Mood-aware Preset moods shift delivery — warm for casual, clinical for a serious note — so tone carries meaning, not just words.
Queue-based Messages don't stomp each other — they wait their turn, so speech stays intelligible under load.
Ambient presence Not an assistant that yells for attention — one that occasionally, quietly, says something.

How it flows

The pipeline is fire-and-forget: the command hands text to a local server, which runs inference and plays the result through the system audio in a detached process — so playback survives the caller exiting. A small routing layer can send the audio to this machine, a paired device, or both, from one entry point.

say "…" local server neural inference detached playback device routing

Under the hood

Built on a GPT-SoVITS voice model (voice-cloned from a custom reference profile), driven by Python, wired into the machine through systemd and a shell entrypoint, and played over the local audio stack. Because synthesis is non-blocking and detached, the caller never waits on it — the request returns in well under a second while the voice plays out on its own.

Python GPT-SoVITS systemd detached playback multi-device routing 100% offline

Where it fits

Vox is the delivery mechanism for Anima's whispers. When Anima detects something worth speaking, Vox gives it a voice — and together they make a background presence: something watching, waiting, and occasionally saying hey.

The say command — one word to a voice, with nothing in between and nothing sent away.