I’m Dave, and I built UTTS—a universal text-to-speech interface that makes large-scale comparisons of TTS models as frictionless as running a single script. Instead of juggling multiple UIs, you can now spin up a Colab notebook or install via pip in seconds.

Why UTTS?

I needed to benchmark dozens of voices and settings across open-source models and commercial APIs, but switching UIs was a huge time sink.

It’s designed for extensibility—drop in new models by implementing a simple interface, then run unified evaluations.

It’s not focused on real-time inference or deep integration with a single model; it’s all about batch comparisons.

Get started

GitHub: https://github.com/arch1baald/utts?tab=readme-ov-file#utts

Colab demo: https://colab.research.google.com/github/arch1baald/utts/blo...

I’d love feedback on new model integrations, metric ideas, UX improvements, or any features you’d like to see. Feel free to comment here or open an issue on GitHub.

Enjoy,

– Dave