Luna AI is a virtual AI streamer that combines state-of-the-art technologies. Its core consists of a series of efficient artificial intelligence models, including ChatterBot, GPT, Claude, langchain, chatglm, text-generation-webui, Xunfei, Zhipu AI, Google Bard, Wenxin Yiyuan, Tongyi Xingchen, Qianfan Big Model, and Gemini. These models can run locally or be supported through cloud services.

The appearance of Luna AI is created using technologies such as Live2D, Vtube Studio, xuniren, and UE5 combined with Audio2Face, providing users with a vivid and interactive virtual image. This allows Luna AI to engage in real-time interactive streaming on major platforms such as Bilibili, TikTok, Kuaishou, WeChat Video Number, Douyu, YouTube, and Twitch. Additionally, it can engage in personalized conversations in a local environment.

To make communication more natural, Luna AI uses advanced natural language processing technology combined with text-to-speech systems such as Edge-TTS, VITS-Fast, elevenlabs, bark-gui, VALL-E-X, RevoiceAI, genshinvoice.top, and tts.ai-lab.top. This not only enables it to generate smooth responses but also allows for voice modulation using so-vits-svc and DDSP-SVC to adapt to different scenarios and roles.

Moreover, Luna AI can collaborate with Stable Diffusion through specific commands to showcase artworks. Users can also customize scripts for Luna AI to play in a loop to meet various needs.

This project is completely free. If an identical program is found being sold as a shell, it is considered pirated, please take timely measures.

▶︎ Video Tutorial Collection | 📄 Online Documentation | 🍉 GitHub | 🍓 Gitee

Thanks to the following developers for their contributions to this project:

Zhilaixingqiu: m.zlxqai.com ———— Provides GPT and AI drawing services

Qingjiaoyun: qingjiaocloud.com ———— AIGC high-performance cloud desktop

RevoiceAI: reecho.ai ———— Instantly creates super-realistic AI cloned voices