AIToolScan

OpenSpeech Overview

OpenSpeech is an open directory of open-source text-to-speech models. Every voice reads the same three standardized scripts so you can hear and compare TTS models side-by-side before choosing one.

• Side-by-Side Comparison: Hear leaders like Kokoro-82M, Orpheus TTS, and Chatterbox back-to-back on identical Neutral, Emotional, and Numbers & Dates scripts.
• Community Rankings: Start with community-voted picks — Best overall, Runner-up, and Third place — updated by votes.
• Browse by Need: Curated shelves for voice cloning, real-time streaming, multilingual, edge deployment, and emotion\/flavor.
• Full Model Directory: Browse 26 open-source TTS models and 46 voices with filters and sort for deeper research.
• Compare Bar & Arena: Add models to a comparison bar and use Directory, Compare, Arena, and Calculator tools to decide faster.
• Standardized Samples: Same three scripts for every voice — baseline naturalness, prosody\/emotion, and number\/date pronunciation.
• Community Contributions: Add a model via a single PR — edit one JSON file, drop three audio samples, open a pull request. Models must be genuinely open-source.
• Open & Independent: Not affiliated with model authors; all models link to original sources. Powered by Gizmo.