Do you recommend any good TTS model for generating audio?

[HarderSoftware]

Hi! Thanks for sharing this great project.

Do you have any recommendations for a text-to-speech model that works well for generating audio in this task?

[xid32]

Hi! Thanks for your kind words and interest in the project.

For text-to-speech (TTS) generation in this task, we’ve found several open-source models to be highly effective. Many thanks to the developers for making these resources available:

1. WhisperSpeech
2. Spark-TTS
3. MegaTTS3 (used in our paper)
4. Zonos

Each of these models offers strong audio quality and flexibility for reasoning-style speech synthesis.