# Inworld 发布 Realtime TTS-2 与 TTS-2 Flash，支持 5-15 秒语音克隆

- 来源：🚨 AI News | TestingCatalog (@testingcatalog)
- 发布时间：2026-09-03 04:08
- AIHOT 分数：51
- AIHOT 链接：https://aihot.virxact.com/items/cmtkjuhq403qxrollwkjnk66c
- 原文链接：https://x.com/testingcatalog/status/2095242569131155858

## AI 摘要

Inworld 发布 Realtime TTS-2（GA，官方称居 Artificial Analysis 榜首）与面向低延迟高并发的 TTS-2 Flash。Realtime TTS-2 可用 5-15 秒音频克隆真实声音，专业级语音克隆处于 beta 阶段需 10 分钟音频，内置 verbatim 标签可正确读出订单号和代码。

## 正文

Inworld has released Realtime TTS-2, focused on clarity, and TTS-2 Flash, built for low latency and high volume.

Inworld Realtime TTS-2 can create a realistic voice clone based on a 5- to 15-second clip.
Professional voice cloning is also in beta and requires 10 minutes of audio input.
A built-in verbatim tag reads order numbers and codes correctly.

### 引用推文

> Inworld AI：Realtime TTS-2 is GA today, now the #1 model on Artificial Analysis and fastest TTS in the world of its class. The research preview got far more usage than expe...