DyVox

DyVox is a text-to-speech application that also acts as a library for users to access already generated audio to reduce the need to unnecessarily spend tokens on any duplicative generations.

• 100% free for commercial use
• Specify speech style (angry, happy, sad, etc)
• 30 high-quality voices
• Combine existing generations
• Google API key slot (for unlimited* generations)
• Fast text-based filtering

GitHub wiki for documentation.

*Using a Google API key of your own bypasses the token system, allowing you to utilize your key. Users who have paid accounts or higher-tiers can basically have near-unlimited generations