Tofu ToFu
Self-managed AI assistant workspace supports multi-model conversations, tool invocations, web searches, project collaboration and automated tasks.
Tofu ToFu is a self-hosted AI assistant workspace deployed on Lazy Cat Micro Clothing. It can be connected to large model services such as DeepSeek, OpenAI, Tongyi Thousand Questions, Gemini, Claude, and OpenRouter for daily question and answer, data search, paper reading, long task disassembly, automated tasks and personal work organization. ##What is suitable for doing - Daily AI dialogue: writing, translating, summarizing, rewriting, brainstorming, code interpretation. - Online search: Let the model search the web first and then answer based on the search results. - Task disassembly: Break complex requirements into multiple rounds of execution, inspection and optimization. - My schedule and to-do: Use my day to organize current tasks and next steps. - Automated tasks: Create tasks such as scheduled reminders, periodic summaries, and data inspections. - Multiple model comparison: Switch between different models in the same conversation and compare the quality and cost of answers. ##First use 1. After opening the app, click `+ New Chat` in the upper left corner to create a new conversation. 2. Click on the gear icon on the left or upper right corner to enter Settings. 3. Add your big model API to Service Provider: - It is recommended to use 'Automatic Configuration' to allow the system to automatically discover the model after filling in the API address and key. - You can also add service providers such as OpenAI, DeepSeek, Tongyi Qianwen, Gemini, and OpenRouter from the template. 4. Select a commonly used model in the Display or Model list and set the default model. 5. Go back to the chat interface, select the model in the model drop-down box at the bottom, enter the question and send it. When the model service provider is not configured, the application interface can be opened, but the AI reply may not be generated normally.

