EdgeSpeak Broadcast

Turn text into speech, entirely on-device — the text never leaves the machine. Broadcast is EdgeSpeak's speech feature; under the hood this skill calls edgespeak-cli speech (alias: synthesize). When the EdgeSpeak desktop app is running, the CLI talks to its local gateway (OpenAI-compatible, 127.0.0.1:1117) and reuses the warm model (proxy mode); when the app is not running, the CLI launches the bundled on-device engine itself (standalone mode). Standalone is a normal mode, not an error.

Version compatibility. The frontmatter pins this skill's version and the oldest CLI it is written against (minCliVersion). If edgespeak-cli --version reports something older, run edgespeak-cli update (or re-run the installer) before relying on the flags documented here. Same-numbered builds can still differ, so --help is the tiebreaker for flags — a flag documented here but missing from the installed --help means update, don't route around it.

--help is not the tiebreaker for model ids: speech --help names only a subset of the installed TTS models. The live, authoritative list is the gateway's /v1/models — see "Pick a model and a voice".

Inputs to confirm

• The text to speak (or the file it comes from).
• Output WAV path.
• Any requested voice, style instructions, speed, language, or reproducibility (seed) preferences.
• Whether the user wants a named voice (a specific, reusable EdgeSpeak voice identity), a cloned voice (their own user: voice), or a designed voice (invented from a text description) — the answer decides the model, not just the --voice value.

How to do it

1. Check the runtime first: