edge-tts Skill (aahl) for Claude Code: What It Does and How to Install It
Makes Claude turn text into an MP3 with Microsoft Edge's neural voices, using the uvx edge-tts command.
- Installs
- 8.4K
- on skills.sh
- Stars
- 165
- on GitHub
- Updated
- Sep 2026
- last repo commit
The skill gives Claude the exact uvx edge-tts commands to write an MP3 file from your text. It also covers speed, volume and pitch, switching voice, and writing subtitles. It lists 33 voices across en-GB, en-HK, en-US, fr-FR, zh-CN, zh-HK and zh-TW, and says to run uvx edge-tts --list-voices for the full list.
It triggers on the keyword "tts", when content should be spoken rather than read, or when you ask for a voice, speed, pitch or format. SKILL.md describes the service as Microsoft Edge's neural TTS service. It bundles no scripts and needs uv installed.
Good for
- People who want an article or summary read aloud while driving, cooking or multitasking
- Anyone who needs spoken output for accessibility
- Creators who want a chosen voice, speed or pitch, plus subtitles
Install edge-tts
Claude Code
- 1
Download the skill folder from https://github.com/aahl/skills/tree/HEAD/skills/edge-tts.
- 2
Put it at ~/.claude/skills/edge-tts/ so that SKILL.md sits directly inside.
- 3
Start a new session; the skill loads when its description matches the task.
Requires uv.
The README documents npx skills add aahl/skills, which installs the whole collection (the repo tree has 15 skill folders). The commands below install only edge-tts. The skill runs uvx, so install uv first and there is no Claude app tab.
Built from the package data we hold. Not yet tested by us.
What's in the skill
When it loads
Text-to-speech conversion using `uvx edge-tts` for generating audio from text. Use when (1) User requests audio/voice output with the "tts" trigger or keyword. (2) Content needs to be spoken rather than read (multitasking, accessibility, driving, cooking). (3) User wants a specific voice, speed, pitch, or format for TTS output.
Sections in SKILL.md
- Usage
- Changing rate(speed), volume and pitch
- Changing the voice
- Available voices
Key facts
- License
- MIT (LICENSE file)
- Installs on skills.sh
- 8,435 (Nov 2, 2026)
- GitHub stars
- 165 for the whole aahl/skills repo (Nov 2, 2026)
- Last repo commit
- Sep 15, 2026
- What it includes
- 1 file: SKILL.md; no scripts
What it includes
uvx edge-tts --text "..." --write-media file.mp3
Writes an MP3 file from the text to a temp directory path the agent chooses.
Try: “tts: read this summary aloud”
Runs code ·
Basic synthesisAdd --write-subtitles - to also output subtitles
Uses the same command with the subtitles flag.
Runs code ·
Subtitles--rate=+50%, --volume=+50% and --pitch=-50Hz
Values shown in SKILL.md are examples.
Runs code ·
Rate, volume and pitch--voice zh-CN-XiaoxiaoNeural and 32 other listed voices
The listed voices are English (GB, HK, US), French and Chinese (CN, HK, TW).
Runs code ·
Change the voiceuvx edge-tts --list-voices
SKILL.md says to use it to see every available voice.
Runs code ·
List all voices
From the official docs
Permissions
Medium risk
What it can do
- Read your data
- Create and change things
- Delete things
- Read-only mode
Watch out for
Your text is sent to an online speech service
SKILL.md describes the tool as using Microsoft Edge's neural TTS service. Do not convert confidential text. github.com
It runs a package through uvx
uvx edge-tts downloads and runs a third-party Python tool. Install uv from its own docs first. github.com
Security advisories
No published security advisories as of Nov 2, 2026.
Technical details
| Talks to | Microsoft Edge neural TTS service (named in SKILL.md) |
|---|---|
| Scopes |
|
Questions
What does the edge-tts skill do?
It makes Claude generate speech audio from text with the uvx edge-tts command. The output is an MP3 file, with optional subtitles and adjustable voice, speed, volume and pitch.
Do I need an API key or an account?
SKILL.md names no key or sign-in. You do need uv, because the commands run through uvx.
Where does my text go?
SKILL.md describes the tool as using Microsoft Edge's neural TTS service, so the text you convert goes to that service. Our advice: do not convert private or confidential text.
How do I install it?
Use the install box above. The README documents npx skills add aahl/skills, which installs the whole collection. The skill needs uvx on your machine.
Which voices and languages are there?
SKILL.md lists 33 voices for English (GB, HK, US), French and Chinese (CN, HK, TW). It says uvx edge-tts --list-voices shows all of them.
Does it work in the Claude app?
The README does not document it. The skill runs uvx on your machine, so the install box has no Claude app tab.
Is it safe?
It runs uvx edge-tts, which fetches and runs a Python package, and sends your text to an online service. Our advice: check what text Claude passes in and where the MP3 is written.
Information
updated Nov 2, 2026
- Repository
- aahl/skills
- Folder
skills/edge-tts
Sources
Docs, issue trackers and discussions this page is built from.