SillyTavern
Provides a unified, locally installed user interface for interacting with text generation large language models, image generation engines, and text-to-speech voice models.
概要
SillyTavern is a robust, locally installed user interface designed to empower AI enthusiasts with unparalleled control and utility over their large language model prompts. Evolving from a 2023 fork of TavernAI, it has become a leading open-source project supported by a dedicated community. It offers a single, comprehensive interface for connecting to diverse LLM APIs (including KoboldAI/CPP, Horde, NovelAI, Ooba, Tabby, OpenAI, OpenRouter, Claude, Mistral), integrating with popular image generation services (like Automatic1111 & ComfyUI), and utilizing TTS models, all while upholding a strict commitment to user privacy and data security.
主な機能
- Extensible architecture supporting a wide array of third-party extensions for features like auto-summary, UI translation, and web search
- Rich UI customization options, a mobile-friendly layout, and a unique Visual Novel Mode
- 15,177 GitHub stars
- Unified interface supporting numerous LLM APIs including OpenAI, OpenRouter, Claude, Mistral, KoboldAI/CPP, and NovelAI
- Seamless integration with Automatic1111 and ComfyUI for image generation, plus multiple Text-to-Speech options
- Advanced prompt management with World Info (lorebooks), user personas, and a powerful scripting engine
ユースケース
- Generating images directly within the chat interface via integrated AI art tools
- Engaging in persistent, customizable conversations with large language models using character cards
- Developing and testing complex LLM prompts with fine-grained control and extensive configuration options
