Omniscio

Voice, TTS & Wake Word (hands-free Omniscio)

Omniscio ships a three-piece voice stack so you can run it without touching keyboard or mouse. Text-to-speech reads agent messages aloud on demand or automatically, dictation is FlowVoice (the OS-wide hotkey types what you say at the cursor, including into Omniscio's own message box), and voice commands turn spoken sentences into actions. A wake word listens on device.

On this page

What it is

Desktop-first, but not desktop-only. Microphone capture, STT, and wake-word processing run inside the Electron desktop app (the AudioWorklet captures audio in the renderer; the main process drives the STT/TTS providers and wake-word engines), so starting a capture is desktop-only. The REST of the stack is reachable without the UI. The localhost CLI control server (127.0.0.1:19519) serves the whole voice family — 23 routes across /voice/*, /voiceprint/* and /flowvoice/*. The two that make sound are approval-gated: POST /voice/speak (speak arbitrary text) and POST /voice/speak-session (speak a session's last message) queue for your approval and transmit only on it. Dictation is driven by POST /flowvoice/start / stop / cancel; PATCH /flowvoice/settings changes its options, POST /flowvoice/recordings/:id/retry and DELETE /flowvoice/recordings/:id handle a saved recording, and GET /flowvoice/state / settings / recordings read them back. Voiceprint exposes GET /voiceprint/profiles / corpus / guide, POST /voiceprint/pull-corpus, POST /voiceprint/corpus/:id/tag, POST /voiceprint/apply-rule and the full-trust-only DELETE /voiceprint/profiles/:id (the right-to-be-forgotten erase — a scoped agent token cannot call it). Reading is GET /voice/status / l4-queue / history / history/:id/turns, and DELETE /voice/history/:id erases one stored voice conversation. A CLI caller can open the mic — POST /flowvoice/start begins a real capture — but that is the CLI source: its transcript comes back on POST /flowvoice/stop for the caller to read, and is never typed at your cursor or submitted as if you had said it. Spoken narration also works in the phone / web client: tap a message's play button and the audio plays on the device, with auto-play of new replies best-effort because mobile browsers block sound until you interact with the page (see spoken-narration.md).

Omniscio ships a three-piece voice stack so you can run it without touching keyboard or mouse. TTS (Text-to-Speech) reads agent messages aloud — either on demand or auto-play as soon as an agent replies. Ten TTS voices ship: Fish Audio (default), Grok (xAI), ElevenLabs, Speechify (Simba), Pika (Pika Speech), Soniox (TTS v2), Gemini TTS, OpenAI (online, with your OpenAI key), and two that run on your own computer with no key and no internet, Piper and Custom command — pick one in Settings → Voice Control → TTS Provider. Dictation is FlowVoice, and only FlowVoice: the OS-wide hold-hotkey (or the composer's mic button) types what you say at the OS cursor, including into Omniscio's own message box. The older in-app dictation — which captured speech and delivered a transcript into the active session's composer — has been retired, so there is no separate in-app dictation left to configure. Voice commands turn spoken sentences into app actions: "open inbox", "answer yes", "create a session in the Api project to refactor the auth middleware", and 40+ more. STT (speech-to-text) runs on Deepgram Nova-3 (~250 ms round-trip), ElevenLabs Scribe v2 (~80 ms), or Groq Whisper v3 Turbo (batch: the full transcript arrives in one piece shortly after you stop speaking, with no live as-you-talk text). Wake word ("Hey Omniscio" or "Computer", or your choice) runs 100% on-device so the mic is always listening but nothing leaves your machine until you say the word — the built-in free engine (openWakeWord) (no key, no account, fully offline; reads model files from a per-user folder) is — as of 2026-07-22 — the ONLY user-selectable engine. The Picovoice Porcupine engine is retired (hidden from the picker, its proprietary models unbundled, its code retained and reversible). Pick in Settings → Voice Control → Wake Word → Detection engine.

Where to find it

Settings → Voice Control holds the master voice stack: the TTS enable toggle, the TTS provider picker with each provider's voice and API-key fields, push-to-talk hold keys, and the Omni Session Briefing controls. The wake word is configured under Settings → Voice Control → Wake Word, and the speech recognition service is picked under Settings → Voice Control. Within the app itself, the microphone button and the Alt+V command hotkey sit with the composer, and Settings → Voice Control → Test voice commands exercises the command parser.

How it behaves

How to use it

  1. Turn on TTS. Settings → Voice Control → flip ttsEnabled. Then pick a provider in TTS Provider (ttsProvider) — Fish Audio (default), Grok (xAI), ElevenLabs, Speechify (Simba), Pika (Pika Speech), Soniox (TTS v2), Gemini TTS, OpenAI, Piper (on this computer), or Custom command (on this computer). Tune volume (ttsVolume); speed (ttsSpeed) applies to Fish only — none of Grok's, ElevenLabs', nor Speechify's API accepts a speed parameter, so the slider is ignored when Grok, ElevenLabs, or Speechify is selected. Flip ttsAutoRead to auto-play every agent reply, or leave it off to play on demand via the speaker button on any message. Long messages are chunked; anything past ttsMaxChars is truncated with an ellipsis.
    • Fish Audio flow. Pick a ttsVoiceId (Fish Audio voice), or add a custom one via Add custom voice → paste a fish.audio voice URL or model ID. A personal Fish Audio key is NOT required — custom voices, like the presets, synthesize and preview through the company relay, so a keyless user can add one, preview it, and hear it in narration. The built-in voices are metered (2026-09-02): each synthesis draws on your plan’s monthly voice time, then on your prepaid credit balance, and when both are empty playback stops with a top-up/upgrade prompt — never a message about an API key you never entered. Free plans include no voice time, so a free account runs entirely on credit. Adding your own key removes the metering entirely: you are billed by Fish Audio directly, with no limit from us. A personal key is optional: paste it in the Fish Audio API Key field and tap Save (Omniscio probes Fish to confirm it) only if you want the live voice-name lookup while adding a custom voice — without a key, Omniscio accepts a well-formed voice ID and confirms it the first time it plays.
    • Grok flow. Pick a grokVoiceId from the five preset voice tiles: Ara, Eve, Leo, Rex, Sal — Eve is the default. Tap a tile to select. Grok tiles ship no bundled preview clips and no per-tile Preview button — you'll first hear the voice when an agent reply auto-plays or when you tap the speaker button on a message. Paste your xAI key in the xAI API Key field and tap Save — Omniscio saves the key and then probes xAI to confirm it works (validation status appears next to the field). Same encrypted-storage UX as the Fish key (safeStorage + enc: prefix in config.json).
    • ElevenLabs flow. Pick an elevenlabsVoiceId (ElevenLabs Voice ID). Paste your ElevenLabs key in the ElevenLabs API Key field and tap Save — Omniscio saves the key and probes ElevenLabs to confirm it works. Like Grok, ElevenLabs accepts no speed parameter, so the speed slider is ignored when ElevenLabs is selected. Same encrypted-storage UX as the Fish key.
    • Speechify (Simba) flow. Paste your Speechify key and tap Save — Omniscio saves the key first, then validates it against GET https://api.sws.speechify.com/v1/voices (validateApiKey('speechify', key)). Once the key is saved, the Speechify Voice field becomes a picker populated live from your own Speechify voice library (fetched from that same /v1/voices endpoint via TTS_LIST_SPEECHIFY_VOICES — only voice metadata crosses IPC, never the key, which stays main-process only), and a Test voice button beside it auditions the selected voice through the normal preview path (TTS_PREVIEW_VOICE, cost logged as tts-preview). Pick a voice (speechifyVoiceId, default alec) or choose Enter a custom voice ID… to type an unlisted one; if the library can't be loaded (no key, offline, or empty) the field falls back to a plain text box so you're never worse off than before. Speechify is user-key-only (no company relay, like Grok / ElevenLabs), so it stays unavailable until you add your own key. The model is pinned to simba-3.0 and, like ElevenLabs, Speechify accepts no speed parameter, so the speed slider is ignored when Speechify is selected. Same encrypted-storage UX as the Fish key (safeStorage + enc: prefix in config.json, main-process only). Note: on the free tier Speechify rate-limits to 1 request/second (a 429 surfaces the friendly "rate-limited — wait a moment" notice).
    • Pika (Pika Speech) flow. Pick a pikaVoiceId from the eight preset voices (Documentary Narrator [default], News Anchor, Radio Alto, and more). Paste your Pika key in the Pika API Key field (from dev.pika.art/keys) and tap Save — Pika is user-key-only (no company relay, like Grok / ElevenLabs / Speechify), so it stays unavailable until you add your own key, and there is no live key validation (Pika has no validation endpoint, so the key is accepted as entered). Unlike the streaming providers, Pika Speech is an async job API — it POSTs the text (script + voice_preset) to api.dev.pika.art, polls until the clip is ready, then downloads the mp3, so a synth takes a few seconds rather than streaming instantly (best suited to narration clips). Like ElevenLabs / Speechify it ignores the speed slider. If Pika errors or runs out of credits, spoken narration auto-falls back to the free Fish voice (the shared out-of-credits safety net) — except in a private AI Coaching or Journal session, which never switches voices and says the recap could not play instead. A private session's recap is in fact never spoken by a company-relayed voice at all: with no key of your own (the default state) the relay is what would answer, so the recap reports that it could not play with your voice rather than using it. Same encrypted-storage UX as the Fish key.
    • Soniox (TTS v2) flow. Pick a sonioxVoiceId from the preset voices (Emma [default], Daniel, Grace, and more) or Enter a custom voice name to type any of Soniox's ~70 named voices. Like Fish, Soniox is company-funded — a personal key is optional: presets and the Test voice button synthesize through the company relay (sonioxTtsRelay) with no key, so a keyless user can pick a voice and hear it (no key is ever required to preview, unlike Speechify/Pika). Like Fish, that keyless path is metered (2026-09-02) — plan voice time, then prepaid credit, then a top-up/upgrade prompt. Add your own key (from console.soniox.com) to bill your OWN Soniox account instead and drop the metering. Add your own key (from console.soniox.com) in the optional Soniox API Key field only if you want to bill your OWN Soniox account instead of the company one. Soniox v2 supports emotion tags ([excited], [calm], [chuckles], …), so spoken narration can direct delivery — the narration prompt teaches the agent the active voice's tags automatically. Like ElevenLabs/Speechify it ignores the speed slider at synth (Soniox speed is applied per request). Same encrypted-storage UX as the Fish key.
    • OpenAI flow. Pick OpenAI in the voice picker, then pick a model (openaiTtsModel: gpt-4o-mini-tts [default], tts-1, or tts-1-hd) and a voice (openaiTtsVoice: alloy [default], ash, ballad, coral, echo, fable, nova, onyx, sage, shimmer, verse). OpenAI uses the same OpenAI API key Codex uses — the key card saves through the OpenAI key channel (openai:set-api-key), so there is one OpenAI key in the app, and it is checked against GET https://api.openai.com/v1/models after saving. OpenAI is user-key-only (no company relay), so it stays unavailable until that key exists. The text is POSTed to https://api.openai.com/v1/audio/speech as mp3, and the speed slider applies. The audio cache is keyed on the model as well as the voice, so changing the model never replays old audio. The metered estimate uses the dearest published rate (tts-1-hd, $30 per million characters) so the daily cap trips early rather than late. OpenAI is already a listed subprocessor; read-aloud and narration text now travels to it when this voice is chosen.
    • Piper and Custom command flows (voices that run on this computer). Pick Piper (on this computer) and choose the Piper program (piperExecutablePath) and a .onnx voice model (piperModelPath) with Browse… or by typing a path. Or pick Custom command (on this computer) and choose any program (customVoiceCommand.executable) plus an editable list of arguments in which exactly ONE argument is the whole token {output} — the card says what is wrong in plain words and offers an Add {output} shortcut. In both cases the text to speak is sent to the program on its standard input, never in its arguments, and the program writes an audio file that Omniscio checks (it must be at least about 1 KB and start like a WAV or an MP3, else the run is refused with a plain sentence). The program is started with exactly the arguments listed and no console window (on Windows a small helper starts it so Omniscio can stop it and anything it started, and the text reaches its standard input through a temporary file that is deleted when the run ends), and is stopped by Stop or after two minutes. Its own error output goes to the log only, with secrets redacted. These voices need no key, cost nothing, and never fall back to an online voice. Omniscio itself sends nothing to the internet for them (a program you choose may do whatever it likes), and a long answer in summarize mode is shortened by a fixed rule-based sentence, never by an online model, so the text never leaves this computer through Omniscio: if one is not ready or fails, nothing is spoken (spoken narration does not retry on Fish, unlike the cloud voices). The audio cache is keyed on the program, model and arguments, so changing any of them gives fresh audio. Preview Voice stays off until the voice is ready. The program and argument settings cannot be changed over the web-access bridge or the CLI.
  2. Try a voice command. Press Alt+V (the dedicated command hotkey), say something like "open the inbox" or "snooze this session for one hour". Release → Omniscio transcribes via your chosen STT provider, parses intent, and runs the matching action. On Mac the Alt+V chord works everywhere EXCEPT while the cursor is in a text field, so typing can never false-trigger it; from a text field use the wake word, the mic button, or rebind Voice command to a Cmd combo. You can also trigger commands via the "Test voice commands" button in Settings → Voice Control. (Note: the toolbar microphone button and the wake word also start a command capture; dictation into the composer is FlowVoice's own mic, not this one.)
  3. Dictate with FlowVoice. Dictation has exactly one home: FlowVoice. Hold its hotkey (Ctrl+Windows by default) or click the FlowVoice mic in the message composer, speak, and release — the transcript is typed at the OS cursor, so it lands in whatever field is focused, including Omniscio's own message box. This replaced the older in-app dictation, which is retired: there is no longer a separate mic that delivers a transcript into a session's composer, and no delivery setting to choose between sending and reviewing first.

    Push-to-talk (hold a key anywhere in the OS, release to stop) ships as one optional system-wide hold key: Push to Command (hold, speak a command, release to run it, with the normal chime + result toast). The former Push to Dictate slot is gone with the in-app dictation it fed — hold-to-talk dictation is FlowVoice's hotkey now. Configure Push to Command in Voice Control settings under Push to talk: it is off by default with no key pre-filled (settings voicePttCommandEnabled/voicePttCommandKey); the picker suggests layout-safe keys (Right Option / Right Command on US Mac layouts, Right Ctrl / Right Shift on AltGr layouts) and Advanced (bind any key) records anything, including bare modifiers. Presses shorter than 100 ms are ignored (accidental-brush guard), and PTT requires the master Enable Voice Commands toggle. macOS: the controls stay locked behind a Grant Accessibility Permission banner until the grant lands (the same grant FlowVoice uses). Right Windows is never suggested on Windows because releasing it would open the Start Menu. If the system key listener dies, the group shows a restart notice (hook supervisor states: ok / restarting / dead). The hold key reuses the same engine, providers, and feedback surfaces as the in-app command mic.

  4. Chain commands. Use "and then" / "then" / "after that" to run multiple in one breath: "archive this session and then open the api project".
  5. Enable wake word for hands-free dictation. Flip wakeWordEnabled, then pick a Detection engine (wakeWordEngine, default openwakeword): the built-in free engine (openwakeword, the default) needs no key and works out of the box with two built-in wake words (Hey Omniscio, Computer) whose model files install automatically on first use into <userData>/wake-word-models/ (the Settings card shows the exact folder, per-keyword model status, and a "Check again" button); Picovoice needs a paid access key (paste it in the key field, which only shows for this engine). The default keyword is Hey Omniscio; change in wakeWordKeywords. Tune wakeWordSensitivity (0.0–1.0, default 0.7 — lower if you get false triggers, higher if it misses you; on the free engine this maps linearly to a score threshold, 0.7 → 0.60). wakeWordFocusOnly (default true) only listens when Omniscio is focused, which saves CPU and feels less creepy. When the wake word fires, the words that follow are parsed as an in-app command (see step 2) — the wake word no longer dictates into a session.

Watch out — "pause" now pauses, not interrupts. Saying a bare "pause" sets the active session's status to paused (mutually exclusive with snooze). It used to be a synonym for interrupt — that wiring was removed. To interrupt the agent mid-turn, say "stop", "interrupt", "cancel", or "halt".

New Tier 1 session controls. "snooze" / "snooze this session" / "snooze for 30 minutes" snoozes the active session (default 60 min; "for N minutes" or "for N hours" is parsed — "snooze for 2 hours" → 120 min). "approve" / "approved" / "yes" / "yeah" / "yep" / "yup" / "sure" / "go ahead" / "do it" / "sounds good" / "looks good" / "permission granted" / "grant permission" all send a fixed affirmative text reply (literally 'Yes, go ahead.') to the active session — it sends a chat message and does not click any approval button or answer a tool-permission prompt; structured in-chat approval is a future capability.

New Tier 2–3 navigation and inbox commands. Three commands were added in the 2026-05-22 voice-control-plane Plan 2. Switch to session — say "switch to the <name> session" or "go to session <name>" to jump directly to a named session using fuzzy name matching. Approve inbox item — say "approve the request", "accept the request", or "approve the cron job" to confirm the currently selected inbox approval (cron job, automation, recipe step, or similar). Reject inbox item — say "reject the request", "decline the cron job", or "deny the request" to dismiss the currently selected inbox approval. Both inbox commands require an approval card to be selected in the Inbox view; if nothing is selected the command is a no-op. Voice snooze of inbox items and reading inbox items aloud are not yet available.

  • notifications-and-silence.md — silence suppresses TTS auto-read the same way it suppresses sounds
  • use-quick-responses.md — Alt+Z / Alt+1 / Alt+2 / Alt+3 are the keyboard parallels to voice replies
  • voice-l4-ask-agent.md - Voice L4 managed ready-queue: ask several running agents at once, hear answers one at a time (go ahead / next / skip / what's ready?)

For the engines and the rest of the stack, see Voice, TTS & Wake Word (part 2), which covers the wake-word engines, the speech-to-text providers, the Omni Session Briefing, spoken output that follows the content's language, third-party dictation and screen readers, and spoken thinking fillers.

Last verified 2026-10-30