Unverified50% confidenceSolutionExact time
应用使用独立小型嵌入模型 all-MiniLM-L6-v2 将信息转化为向量并存储在不绑定任何聊天模型的本地索引中
1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen
Sources
大学生用Ollama自建跨模型持久记忆系统：解耦设计+AES256加密
Related Claims
Unverified该工具提供Token用量查看功能，用户需自行配置对话模型和生图模型68% similarUnverifiedThe project uses Ollama for chat and embedding models, ChromaDB as the vector store, and Flask as the web framework, implementing a fully offline PDF question-answering system.65% similarUnverifiedThe platform supports cross-model conversation context retention, meaning that after switching models, previous conversation records are still preserved.64% similarUnverifiedAnythingLLM 允许将所有数据和模型推理保留在本地设备上，数据从输入到输出全程不离开用户设备63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/503463API
curl https://kongchang.com/api/v1/knowledge/claims/503463MCP
get_claim(id=503463)