Show HN: Voxica – Multi-character AI conversations with ambient sounds

Existing TTS tools usually support only single voices, but I needed natural multi-character conversations for audiobooks and podcasts—so I built Voxica.

Live demo: https://voxica.io (try the "Maya & Noah" conversation)

Voxica generates:

Multi-character dialogue from simple text input

Automatic voice switching

Realistic background sounds (café, rain, office, etc.)

Export-ready audio files

Perfect for:

Podcast/audiobook creators

Developers building voice apps

Educators making interactive lessons

Anyone who needs conversation audio quickly

I built this after spending weeks manually editing podcast voice clips—now it takes 30 seconds instead of 3 hours.

API docs: https://voxica.io/docs Free to try, usage-based pricing after that.

Curious what other use cases or conversation scenarios people here would find useful!