⌘K
Change language Switch ThemeSign In
Narrow Mode
Fish Audio S2: Next-Gen High-Performance Open-Source TTS Model Released =======================================================================
Fish Audio S2: Next-Gen High-Performance Open-Source TTS Model Released =======================================================================  ### AIGCLINK
@aigclink
Fish的TTS刚出了新款：Fish Audio S2，RTF 0.195，首包延迟100ms，单次生成可包含多个说话人
音质清晰和表达自然度在TTS里算是可以的
支持80+语言、长文本、15000+种自然语言描述标签进行语音控制
完全可以应对实时对话、多角色故事、长文本朗读的场景
#TTS #FishAudioS2
00:49
#### Fish Audio
@FishAudio · 11h ago
Today we launch Fish Audio S2, a new generation of expressive TTS with absurdly controllable emotion.
- open-source
- sub 150ms latency
- multi-speaker in one pass
00:49
89
158
1,237
1M
Mar 11, 2026, 1:55 AM View on X
1 Replies
2 Retweets
11 Likes
3,325 Views  AIGCLINK @aigclink
One Sentence Summary
Fish Audio launches the S2 version of its TTS model, featuring ultra-low latency, multi-speaker support, and advanced voice control capabilities.
Summary
This tweet details the core features of Fish Audio S2, a new text-to-speech (TTS) model. The model delivers outstanding performance with a Real-Time Factor (RTF) of 0.195 and first-packet latency of only 100ms, while supporting multiple speakers in a single generation. Technically, it supports over 80 languages and allows for precise voice control via 15,000+ natural language description tags. As an open-source model, it is perfectly suited for scenarios requiring high real-time performance and expressiveness, such as live dialogue, multi-character storytelling, and long-form narration.
AI Score
84
Influence Score 5
Published At Today
Language
Chinese
Tags
TTS
Fish Audio S2
Speech Synthesis
Open Source Model
AI HomeArticlesPodcastsVideosTweets
Fish Audio S2: Next-Gen High-Performance Open-Source TTS ... ===============