Soul plans to launch Loxinity for real-time AI audio-video interaction
According to reports from 《读佳》, Soul plans to launch Loxinity, an AI audio-video product for interactive scenarios. The product combines the Loxi Speech speech-synthesis model and Loxi Avatar video-generation model, with modules for speech synthesis, real-time video calls and digital humans. It supports single-person dubbing, multi-person conversations, voice cloning and design, API access, and customizable digital-human images. The reports say the product emphasizes conversational details such as pauses, intonation and emotional variation, and uses both credits and subscription plans. Its commercial performance remains uncertain because long-conversation stability, concurrency, end-to-end latency, project costs and digital-human naturalness require further validation in external B2B deployments.
Loxinity combines speech synthesis, real-time video calls, digital humans and API access in one interactive product, but its external B2B performance and market readiness have not been established.
《读佳》独家获知，Soul将推出专为交互场景打造、超具活人感的语音合成产品“Loxinity（洛希无限）”，和市面上面向配音、短视频的TTS不同，这套模型从底层瞄准实时交互场景，主打强活人感。它支持单人配音、多人对话两种模式，配套音色克隆、音色设计等工具，还开放API接入。Loxinity下辖语音合成模型Loxi Speech和视频生成模型Loxi Avatar。
目前看，比起只追求字音标准的传统语音模型，Loxinity更看重对话里的语气停顿、情绪起伏这类副语言细节，用来承接社交场景里的连续语音互动。
该产品的使命为：让 AI 交互更自然——用语音、视觉与多模态技术缩短人与智能之间的距离。逼格不低，不排除这款产品成为Soul在社交业务之外，在AI语音场景的一次重要尝试。
Soul把AI语音和实时视频通话、数字人打包推向市场
《读佳》实测Soul旗下AI产品Loxinity发现，这套产品基于语音合成模型Loxi Speech和视频生成模型Loxi Avatar。而且不只是单一语音合成工具，而是一套面向实时交互的AI音视频一体化方案，核心分为语音合成、实时视频通话与数字人三大板块。