Hi HN,<p>I just launched TTS Arena V2 - a platform for benchmarking TTS models by blind A/B testing. The goal is to make it easy to compare quality between open-source and commercial models, including conversational ones.<p>What's new in V2:<p>- Conversational Arena: Evaluate models like CSM-1B, Dia 1.6B, and PlayDialog in multi-turn settings
- Personal Leaderboard: Optional login to see which models you tend to prefer
- Multi-speaker TTS: Random voices per generation to reduce speaker bias
- Performance Upgrade: Rebuilt from Gradio -> Flask. Much faster with fewer failed generations.
- Keyboard Shortcuts: Vote entirely via keyboard<p>Also added models like MegaTTS 3, Cartesia Sonic, and ElevenLabs' full lineup.<p>I'd love any feedback, feature suggestions, or ideas for models to include.
Can Dia not generate non-conversational TTS? I also think it will lose in conversational most times because by default it speaks too fast.<p>I like your UI and it feels snappier than the other TTS Arena I have seen. I would still like it to track the personal leaderboard though without logging in.