Tag
Fish Audio 发布了 Drama 3 模型的权重，支持用自然语言指导台词演绎并可在句中切换表演方式，引发关于开源语音模型是否将迎来“DeepSeek 时刻”以及 ElevenLabs 等闭源模型护城河的讨论。
Fish Audio 发布了开源权重的 TTS 模型 Drama 3，号称全球最具情感表现力和可控性的语音合成模型，支持在句子中对单个单词进行风格和语气微调。
Fish Audio 推出了 Drama 3 语音模型，支持音频修复（audio inpainting），可以只重新生成音频中的单个词或短语，其余部分保持不变，并支持 83+ 种语言。
The author shares the complete process of integrating the Fish Audio speech service into Muse, with support for Chinese dubbing, multiple voice timbre options, voice cloning, and the ability to call the free s2.1-pro-free model.
Fish Audio 发布了 Drama 3 语音模型，支持在普通语言指令下控制语气、情绪与节奏，甚至能在同一句话中切换情绪，展现出接近人类语音的细腻控制力，现已通过 API 以 'drama-3-preview' 提供预览。
Fish Audio 发布新一代文本转语音模型 Drama 3，采用 35B 参数 MoE 架构（每 token 激活 3B），支持最多 16 个说话人的多角色长文本生成、语音克隆与自然语言控制，并可在 SGLang-Omni 上流式部署。
Fish Audio has made its S2.1 Pro voice cloning service free for a month, featuring 10-15 second voice cloning, ~90ms response time, support for 83 languages, word-level control, and open-weight models at 1/6th the cost of ElevenLabs.
Fish Audio launches S2.1 Pro, a production voice model with 90ms latency, support for 83 languages, voice cloning from short samples, and multi-speaker dialogue, available via API with a free tier for development.
Fish Audio S2 is a new open-weight audio model available on Hugging Face, offering two models for timing and acoustic details, with fast inference and a hosted version S2.1 Pro supporting 83 languages at lower cost than ElevenLabs.
Fish Audio S2 is an open-source text-to-speech system featuring multi-speaker capabilities, multi-turn generation, and instruction-following control, backed by a production-ready inference engine with low latency.