How do I install the Text To Speech skill?

Run npx skillmds@latest add seanx10m/text-to-speech in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Text To Speech safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Text To Speech?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Text To Speech free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Text To Speech?

seanx10m (@seanx10m) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Text To Speech by seanx10m — Agent Skill

Name: Text To Speech
Author: seanx10m

Text To Speech

Convert a blob of text into spoken audio using OpenAI's text-to-speech (a generic ChatGPT-style voice) and hand the audio file back. Use when the user gives text and asks to "read this aloud", "make audio of this", "text to speech", "TTS this", "narrate this", "turn this into an mp3/voiceover", or invokes /text-to-speech. Produces an audio file (mp3 by default) saved to the Desktop.

seanx10m Updated Sep 17, 2026

File contents

text-to-speech

Turns text into an audio file via OpenAI's speech API. The key lives in the macOS Keychain (service openai-api-key, account openai-tts) — the script reads it automatically, so nothing is hardcoded.

How to run it

python3 ~/.claude/skills/text-to-speech/scripts/tts.py "TEXT HERE"

The script prints the saved file path on stdout (one line). After it returns:

Surface the file to the user with SendUserFile so they can grab it.
Offer to play it, or pass --play to play it immediately via afplay.

Input options (pick one)

Inline: tts.py "Hello world"
From a file: tts.py -f /path/to/notes.txt
From stdin (best for large blobs — avoids shell-quoting issues):
```
cat notes.txt | python3 ~/.claude/skills/text-to-speech/scripts/tts.py --play
```
When the text is long, has quotes/newlines, or comes from another file, prefer writing it to a temp file and using -f, or piping via stdin.

Common flags

Flag	Default	Notes
`-v, --voice`	`alloy`	alloy, ash, ballad, coral, echo, fable, onyx, nova, sage, shimmer, verse
`-m, --model`	`gpt-4o-mini-tts`	also `tts-1` (fast) / `tts-1-hd` (higher quality)
`--format`	`mp3`	mp3, opus, aac, flac, wav, pcm
`-i, --instructions`	—	tone/style steer, e.g. `-i "warm, slow, like a bedtime story"` (gpt-4o-mini-tts only)
`--speed`	`1.0`	0.25–4.0; best on tts-1 / tts-1-hd
`-o, --out`	`~/Desktop/tts-<timestamp>.mp3`	output path
`--play`	off	play immediately after saving

alloy is the plain, neutral "generic ChatGPT" voice — the default. Only change it if the user asks for a different tone.

Notes

Standard library only — no pip install needed.
Key precedence: OPENAI_API_KEY env var, then Keychain. Rotate the stored key with: security add-generic-password -a openai-tts -s openai-api-key -w 'sk-...' -U
Errors (bad key, unknown voice/model) are printed as ERROR <code>: <detail>.

---name: text-to-speechdescription: Convert a blob of text into spoken audio using OpenAI's text-to-speech (a generic ChatGPT-style voice) and hand the audio file back. Use when the user gives text and asks to "read this aloud", "make audio of this", "text to speech", "TTS this", "narrate this", "turn this into an mp3/voiceover", or invokes /text-to-speech. Produces an audio file (mp3 by default) saved to the Desktop.---# text-to-speechTurns text into an audio file via OpenAI's speech API. The key lives in themacOS Keychain (service `openai-api-key`, account `openai-tts`) — the scriptreads it automatically, so nothing is hardcoded.## How to run it```bashpython3 ~/.claude/skills/text-to-speech/scripts/tts.py "TEXT HERE"```The script prints the saved file path on stdout (one line). After it returns:1. Surface the file to the user with `SendUserFile` so they can grab it.2. Offer to play it, or pass `--play` to play it immediately via `afplay`.### Input options (pick one)- Inline: `tts.py "Hello world"`- From a file: `tts.py -f /path/to/notes.txt`- From stdin (best for large blobs — avoids shell-quoting issues): ```bash cat notes.txt | python3 ~/.claude/skills/text-to-speech/scripts/tts.py --play ``` When the text is long, has quotes/newlines, or comes from another file, prefer writing it to a temp file and using `-f`, or piping via stdin.### Common flags| Flag | Default | Notes ||---|---|---|| `-v, --voice` | `alloy` | alloy, ash, ballad, coral, echo, fable, onyx, nova, sage, shimmer, verse || `-m, --model` | `gpt-4o-mini-tts` | also `tts-1` (fast) / `tts-1-hd` (higher quality) || `--format` | `mp3` | mp3, opus, aac, flac, wav, pcm || `-i, --instructions` | — | tone/style steer, e.g. `-i "warm, slow, like a bedtime story"` (gpt-4o-mini-tts only) || `--speed` | `1.0` | 0.25–4.0; best on tts-1 / tts-1-hd || `-o, --out` | `~/Desktop/tts-<timestamp>.mp3` | output path || `--play` | off | play immediately after saving |`alloy` is the plain, neutral "generic ChatGPT" voice — the default. Only changeit if the user asks for a different tone.## Notes- Standard library only — no pip install needed.- Key precedence: `OPENAI_API_KEY` env var, then Keychain. Rotate the stored key with: `security add-generic-password -a openai-tts -s openai-api-key -w 'sk-...' -U`- Errors (bad key, unknown voice/model) are printed as `ERROR <code>: <detail>`.

Frequently asked questions