Old Zhang's AI Learning

Feb 7, 2026 · Artificial Intelligence

Zero‑Shot Voice Cloning with Emotion and Duration Control: IndexTTS‑2 Runs Locally

IndexTTS‑2, an open‑source zero‑shot TTS system from B‑Station, enables precise duration control, emotion‑tone separation, and bilingual synthesis, offering a modern uv‑based installation, GPU‑accelerated inference, and benchmark‑leading WER and emotional similarity scores compared to contemporary models.