sapi-tts_skill

This skill provides Windows SAPI5 text-to-speech using neural and legacy voices, with zero GPU usage and instant audio generation.

Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill sapi-tts

_meta.json269 B
SKILL.md6.3 KB

Overview

This skill provides lightweight Windows SAPI5 text-to-speech using built-in voices, including Windows 11 Neural voices. It produces WAV audio instantly with zero GPU usage and auto-selects the best available voice for the requested language. Works on Windows 10 and 11 and includes options to list voices, set language, rate, and play output immediately.

How this skill works

The script loads the System.Speech SAPI5 synthesizer, detects installed and enabled voices, and selects the best voice by language or neural availability. It can save speech to a timestamped WAV file, optionally play it using the .NET MediaPlayer, and reports the chosen voice and output path. Voice selection supports partial name matches, language shortcuts, and a fallback order (neural by language → legacy by language → any neural → first available).

When to use it

Quickly generate TTS audio on Windows without GPU or external services
Embed simple TTS into automation scripts and chatbots running on Windows
Produce instant voice previews during content creation or voice testing
Generate WAV files for accessibility, notifications, or local demos
Play speech immediately for human review or demonstration

Best practices

Use -ListVoices first to discover installed neural or legacy voices before specifying a name
Prefer language codes (en, fr, de, es, it) to let the skill auto-select the best voice
Install additional Neural voices via Windows Settings or a SAPI adapter if you need higher quality voices
Set -Output to a controlled folder if you need deterministic filenames or avoid clutter
Use -Rate sparingly; small adjustments (±1–3) improve pacing without sounding unnatural

Example use cases

Convert short prompts to WAV for local voice assistants or offline demos
Run within automation pipelines to generate spoken notifications or instructions
Record voiceovers for UI prototypes using built-in Neural voices on Windows 11
Quickly test and compare installed voices with -ListVoices then generate sample audio
Create accessibility reads for desktop apps without relying on cloud TTS

FAQ

Windows 10 and Windows 11 are supported. Neural voices are native to Windows 11; Windows 10 can use legacy voices or third-party adapters.

How do I choose a specific voice?

Use -VoiceName (or -Voice / -v) with a partial or full name. If not found the script auto-selects the best match for your language.

Can I get higher-quality voices?

Install additional Neural voices via Settings → Time & Language → Speech or use a Natural Voice SAPI adapter to add more neural voices.