Speekify

Speekify turns text into a local WAV file you can listen to — from inline CLI text, stdin, local .txt/.md/.pdf files, YouTube transcripts, or the readable content of a URL. Audio is generated locally with Supertonic v3.

Features

Install

speekify setup (run once after installing) downloads and warms the Supertonic model and, by default, the English→French translation model. If you skip it, the models download automatically on first use.

macOS — no Python or uv required

Once a release is published, install the standalone binary via Homebrew:

brew tap otterlyspacelabs/speekify https://github.com/OtterlySpaceLabs/speekify
brew install speekify
speekify setup

Or download the archive directly:

curl -L -o speekify.tar.gz https://github.com/OtterlySpaceLabs/speekify/releases/latest/download/speekify-macos-arm64.tar.gz
tar -xzf speekify.tar.gz
mv speekify /usr/local/bin/speekify
speekify setup

With pip / pipx / uv

Once released on PyPI:

pip install speekify        # or: pipx install speekify  (recommended for CLIs)
speekify setup

uv tool install speekify    # uv users
speekify setup

Run without installing:

uvx speekify "Hello world"
pipx run speekify "Hello world"

From source with uv

uv sync
uv run speekify setup

Quick start

Speekify is a CLI. Give it inline text, a URL, a local file path, or piped stdin:

speekify "Hello world"                                        # inline text
speekify https://example.com/article                          # readable URL
speekify ~/Documents/article.pdf                              # local PDF
speekify --lang fr "https://www.youtube.com/watch?v=eSP7PLTXNy8"   # YouTube
printf 'Hello from stdin' | speekify                          # stdin
speekify --voice F2 --output-dir ~/Desktop "Hello world"      # pick voice + dir

Wrap URLs in quotes when they contain ? or & (common with YouTube), or the shell may reject the command. From a source checkout, prefix any command with uv run (e.g. uv run speekify "Hello world").

→ Full command list, every option, and the config file: docs/usage.md.

MCP automation server

Speekify ships a local Model Context Protocol (MCP) server so AI assistants can call it as a tool. It exposes speekify_generate_wav (text/URL/file → WAV with structured details), speekify_generation_defaults (supported voices, languages, and ranges), and a news_recap_to_audio prompt template.

speekify mcp                              # stdio transport for desktop AI clients
speekify mcp --transport streamable-http  # HTTP endpoint for local experiments

Per-client setup (Claude Code, GitHub Copilot, Codex, OpenAI remote MCP): docs/mcp-clients.md.

Documentation

Usage and content

Speekify only fetches content you explicitly point it at, for your own personal use. You are responsible for respecting the terms of service and copyright of any source you extract from, including YouTube transcripts. The maintainers provide no rights over third-party content.

License

MIT — see LICENSE. Copyright © 2026 Otterly Space SARL.