Speak TTS
On-device Apple Silicon TTS CLI with voice cloning, auto-chunking and resume for long documents.
Test report
- Verdict
- Works with setup
- Score
- Tested
- Jul 16, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 50bdfbb
Code-audited rather than executed (no Bun runtime and no throwaway 500MB model download in this sandbox): every documented flag, command, and even exact error string ('Server not running', file-not-found, sox-missing) was found genuinely implemented across a real 2,827-line TypeScript + 752-line Python codebase, not vaporware. But SKILL.md's own Prerequisites table never mentions Bun or Python 3.10+, and the project's own recommended agent quick-setup (copy SKILL.md only) ships a skill that references a `speak` binary that isn't actually on PATH.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 3/5
- Triggers reliably 5/5
- Output vs. baseline 6/10
- Docs & honesty 3/5
What Speak TTS does
Gives an agent a local `speak` CLI (Bun + Python/MLX Chatterbox backend) to read text, Markdown, or piped content aloud, clone a voice from a short sample, batch-process chapters, and resume interrupted long-document generation, entirely on-device with no API keys. Triggers on requests to read something aloud, narrate an audiobook, or give a spoken response, but only functions on Apple Silicon Macs.
How to install Speak TTS
git clone https://github.com/EmZod/speak.git
mkdir -p ~/.claude/skills
cp -r speak ~/.claude/skills/speak-tts
cd ~/.claude/skills/speak-tts && bun install && alias speak="bun run $(pwd)/src/index.ts"
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Speak TTS
-
/speak-ttsOn-device Apple Silicon TTS CLI with voice cloning, auto-chunking and resume for long documents.
It also activates on plain-language prompts like these:
-
Read this article aloud to me using my own cloned voice, offline. -
Narrate this long ebook chapter and let it resume if interrupted. -
Turn this markdown doc into spoken audio without any cloud API keys.
Frequently asked questions
- Is the Speak TTS skill free?
- Yes. The skill itself is free from EmZod/speak. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Speak TTS work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 16, 2026. Verdict: Works with setup. Code-audited rather than executed (no Bun runtime and no throwaway 500MB model download in this sandbox): every documented flag, command, and even exact error string ('Server not running', file-not-found, sox-missing) was found genuinely implemented across a real 2,827-line TypeScript + 752-line Python codebase, not vaporware. But SKILL.md's own Prerequisites table never mentions Bun or Python 3.10+, and the project's own recommended agent quick-setup (copy SKILL.md only) ships a skill that references a `speak` binary that isn't actually on PATH.
- What is the Speak TTS SkillProof Score?
- 6.8/10 — installs cleanly 3/5, triggers reliably 5/5, output vs. baseline 6/10, docs & honesty 3/5.
- How do I install Speak TTS?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Speak TTS with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.