Spaces:

---

Duplicated from  TTS-AGI/TTS-Arena-V2

Pendrokar

/

TTS-Spaces-Arena

Like 493

Running

Adds LinnetTTS (linnet-en-35m, by VoxRT) to the arena.

• Space: https://huggingface.co/spaces/VoxRT/linnet-tts (Gradio, /tts endpoint: text, voice_name -> audio; default voice linnet-en-f1)
• 41M-parameter on-device English TTS, 24 kHz, runs on the Space's cpu-basic in real time (no GPU / ZeroGPU needed)
• Model card: https://huggingface.co/VoxRT/linnet-en-35m, code: https://github.com/VoxRT/LinnetTTS (Apache-2.0; weights under the Linnet Weights License, VoxRT authorises arena evaluation)

Files: one line in app/models.py (AVAILABLE_MODELS), app/tts_spaces/VoxRT__linnet-tts.json, app/inputs/VoxRT__linnet-tts.json.