Do they sound like robots?

No. We use state-of-the-art TTS models (like ElevenLabs) that include natural pauses, breaths, and emotions.