tts · noizai — Claude Code skill: text-to-speech setup
Turns text and documents into speech audio, with subtitle-timed output and voice references. Use it for narration, dubbing, or spoken versions of written material.
npx skills add NoizAI/skills --skill tts -a claude-codeWhat tts does
This skill generates speech from text or files through local Kokoro or the Noiz cloud backend. It can process text, EPUB, PDF, and SRT inputs, write MP3, OGG, or Opus output, and use a reference recording for voice cloning.
For timed audio, it creates an SRT file, applies a JSON voice map to individual segments or ranges, then renders the track. ffmpeg is used only for render, while the Noiz backend requires the requests package. Kokoro runs locally without network access.
When no API key is configured, the script switches to Noiz guest mode with voice, speed, and format controls. It excludes voice cloning, emotion controls, segment duration, and timeline rendering. Text and optional reference audio reach Noiz only when you run its backend, so review the material you send to the cloud.
What it helps with
- turn written text into audio
- make spoken EPUB, PDF, or SRT files
- clone a voice from a reference recording
- align speech with subtitle segments
- assign voices to selected segments
Example prompts for the agent
- Read
article.txtaloud and save the result asout.mp3. - Create
article.srtfromarticle.txt, then render timed audio from it. - Generate audio using
./ref.wavas the voice reference and save it toclone.wav.
How to install tts
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill tts -a claude-code2Manually: put the files in place
~/.claude/skills/tts/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/tts/
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill tts -a codex2Manually: put the files in place
.agents/skills/tts/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill tts -a cursor2Manually: put the files in place
.cursor/skills/tts/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install NoizAI/skills tts2Manually: put the files in place
.github/skills/tts/SKILL.md1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill tts -a gemini-cli2Platform CLI
gemini skills install https://github.com/NoizAI/skills --path skills/ttsManually: put the files in place
.devin/skills/tts/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add NoizAI/skills --skill tts -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about tts
How do I install tts in Claude Code?
Run npx skills add NoizAI/skills --skill tts -a claude-code, then run npx skills add NoizAI/skills --skill tts -a claude-code again. The skill file is at ~/.claude/skills/tts/SKILL.md.
What does tts do?
It produces speech audio from text using Kokoro or Noiz. The skill supports direct narration, reference audio, and SRT-based timed rendering.
What works without an API key?
Noiz guest mode starts automatically and supports voice selection, speed, and output format. It does not include voice cloning, emotion control, or SRT timeline rendering.
Similar skills
Everything for AI and ML →tts by other authors comes first.
