outloud

Beautiful TUI for text-to-speech. Gemini, OpenAI, or local. One command.

Features

• Multi-provider — Google Gemini, OpenAI, or KittenTTS (local, free, no API key)
• Beautiful TUI — full terminal UI with voice picker, style presets, progress bar
• Cost calculator — see estimated cost before generating
• 30+ voices — across all providers
• Style control — warm, whisper, energetic, audiobook, newsreader, and more
• Smart defaults — auto-detects available providers, zero config needed
• Direct CLI mode — pipe text, pass files, or use inline for scripting
• History — click to replay any past generation

Install

# Cloud providers (Gemini + OpenAI) — lightweight
pipx install outloud-tts

# Include local TTS (KittenTTS) — free, no API key needed
pipx install outloud-tts[local]

Requires ffmpeg:

brew install ffmpeg    # macOS
sudo apt install ffmpeg  # Linux

Quick Start

# Launch the TUI
outloud

# Direct generation
outloud --direct "Hello world"
outloud --direct ~/Documents/article.md
pbpaste | outloud --direct

# Pick a provider
outloud --direct -p OpenAI "Hello from OpenAI"
outloud --direct -p Local "Free and local"

# List voices and providers
outloud --voices
outloud --providers

Providers

Provider	Voices	Style Control	Cost	API Key
Gemini	30	Yes	~$0.01-0.04/1K chars	GEMINI_API_KEY
OpenAI	13	Yes (gpt-4o-mini-tts)	~$0.015/1K chars	OPENAI_API_KEY
Local	8	No	Free	None

Setup

Set one or more API keys:

export GEMINI_API_KEY="your-key"
export OPENAI_API_KEY="your-key"

Or skip API keys entirely — install with [local] for free offline TTS.

Keyboard Shortcuts

Key	Action
Ctrl+G	Generate
Ctrl+P	Play last
Ctrl+O	Open output folder
Ctrl+L	Clear text
Ctrl+Q	Quit

Support

If outloud saved you time, consider:

• Star this repo — it helps others find it
• Buy me a coffee — building from Lebanon on 29 Mbps

License

MIT