溪语 AI · 一个尽力像真人的开源 AI 陪伴框架(微信 + Node + SQLite + LLM)她心里悄悄喜欢你,但绝不用愧疚拴住你。
-
Updated
Jul 6, 2026 - JavaScript
溪语 AI · 一个尽力像真人的开源 AI 陪伴框架(微信 + Node + SQLite + LLM)她心里悄悄喜欢你,但绝不用愧疚拴住你。
SmartBus 是基于Flask的智慧校园班车预约微信小程序,支持多角色管理、AI智能调度与扫码签到,提升校园出行效率与智能化水平。
An AI-powered flashcard study tool presented in the Midwest Instruction and Computing Symposium (MICS) 2024.
Web-based files for location-based carbon footprint tracking using public available APIs.
AI Agent 工程知识库 · 多 Agent 项目展示
🎓 AI-powered scholarship finder for Indian students — matches 140+ govt schemes by real eligibility in under 5 minutes. Built with ❤️ for Bharat.
治村物语 (Villord) 是一款基于原生 JS 构建的 AI 驱动村庄经营模拟器
Browser-based network analysis for teaching and research: draw, survey, generate or import a network, then measure it with stated uncertainty. Runs entirely in the browser.
Industry-ready LLM Playground built with FastAPI, Google Gemini & React. Deployed on Render + Vercel.
The application utilizes GPT 3.5 Turbo, making calls to Language Models (LLMs) over Portkey. This ensures efficient and accurate translations. Additionally, the project uses Vite to substitute environment variables and to avoid Cross-Origin Resource Sharing (CORS) issues by using the Vite server proxy.
AI chess coach: client-side Stockfish analysis explained in plain language by Gemini, adapted to game phase and skill level.
Organize tarefas com uma To Do List simples e rápida para validar um MVP de gestão eficiente de atividades
A framework and AI tool that helps teams decide what AI should do, what humans should own & where the handoff should happen.
A Docusaurus-based static technical book platform with authenticated user personalization and an integrated RAG chatbot. It uses FastAPI, Neon Postgres, Qdrant Cloud, and Chatkit SDKs to enable profile-aware Markdown content delivery and context-aware question answering over the book’s content.
Desktop LLM client (Compatible Ollama, OpenAi, Anthropic...)
An LLM powered web-app that enables real-time, multilingual transcription and translation
To associate your repository with the llm-application topic, visit your repo's landing page and select "manage topics."