Edge0_AI's open-weight Audio8-TTS tops Hugging Face TTS Arena at 601M params
realmrfakename · x · 2026-10-30
Edge0AI released Audio8-TTS-Realistic, an open-weight speech model offering realistic zero-shot voice cloning from a single short clip in 23 languages, with human-like prosody.
At 601M parameters it's small enough to fine-tune and serve cheaply, and it currently ranks #1 on the Hugging Face TTS Arena blind-voting leaderboard. Weights are available on HF.