sag
Modern text-to-speech CLI powered by ElevenLabs
Free Text to Speech Open Source
Overview
sag is a command-line text-to-speech tool that works like macOS's built-in `say` command but uses ElevenLabs' advanced voice synthesis for natural-sounding speech. Stream audio directly to speakers, save to files, or pipe text for instant playback. With support for multiple voice models including the highly expressive v3, adjustable speed controls, and voice parameter tuning, sag brings professional-quality TTS to your terminal workflow.
Pricing: Free (requires ElevenLabs API key)
Architecture: Apple Silicon, Intel
Key Features
- Stream TTS audio directly to speakers like macOS say
- Save audio to files with automatic format detection
- Multiple ElevenLabs models: v3 (expressive), multilingual v2, flash/turbo
- Voice discovery and listing with sag voices
- Adjustable speed/rate controls mapped to words per minute
- Voice parameter tuning: stability, similarity, style, speaker boost
- Text normalization and language specification
- Seed-based reproducibility for consistent output
- Install via Homebrew or Go toolchain
- Works on macOS, Linux, and Windows
