MOSS

MOSS-TTS-v1.5 just reached #1 on Hugging Face Trending for Text-to-Speech, with 20.6K downloads. A multilingual, controllable TTS model with stable voice cloning, long-form generation, and precise pause control. MOSS-TTS-v1.5 is now officially supported by vLLM-Omni and SGLang-Omni. Built by Open

Perps are now live on Prism Powered by @GMX_IO, no added fees on top Try MOSS, now in private beta, for a smoother trading experience. Step 1: Click on MOSS Connect (you must be logged out of any wallet) Step 2: Create an account and fund it Step 3: Enable one-click for the duration of your choice

发现一个很良心的开源语音项目 MOSS-TTS 来自 https://t.co/BLkLJ85GIS 和 OpenMOSS 团队，不是单个模型，是一整套语音生产线 能做什么： - 零样本声音克隆，稳得住长文本 - 多人对话生成，主观评分超过豆包和 Gemini 2.5 - 文字描述就能创造新音色，不要参考音频 - 实时流式，首字延迟 180ms - 音效生成，自然声城市声动作声都可以 部署友好这点真不错 8B 模型 llama.cpp 推理，8G 显存够用 还有个 Nano 版，100M 参数，4 核 CPU 就能跑 做配音、虚拟主播、语音助手的朋友值得一试 https://t.co