Local-first GenAI lab with a React frontend and Spring Boot backend, supporting Ollama, Amazon Bedrock, Hugging Face, RAG document retrieval, MCP-backed AWS tools, streaming chat, persistent sessions, and artifact/report workflows.

A modern, free AI chatbot application built with React, TypeScript, and Puter AI services with real-time streaming responses

macOS and iOS clients for self-hosted Hermes Agent API Server: streaming chat, Responses, Runs, tool progress, and Cron Jobs

A lightweight, privacy-focused web interface for interacting with OpenAI-compatible APIs. This single-file application provides a feature-rich chat experience with complete client-side data management, ensuring your conversations and API credentials never leave your device.

自托管 HTTP 编程 Agent 工作台：支持流式对话、持久会话、文件交付、后台任务和 Docker Chrome 真实浏览器访问。

Next.js 16 boilerplate with a pluggable AI provider registry, streaming chat UI, and OpenAI integration out of the box.

A modern conversational AI interface built with Next.js and TypeScript. It integrates with LLM APIs to deliver real‑time, context‑aware chat responses with streaming output.

A fully local AI chat assistant with RAG, semantic memory, and web search. Built with FastAPI + React + Ollama + ChromaDB. Features streaming SSE chat, multi-model support (Ollama / OpenAI-compatible APIs), file upload Q&A, HyDE + BM25 re-ranking, and model warmup. No cloud API required.

Local-first AI console built with Next.js, TypeScript, and Ollama. Includes streaming chat, Quick/Daily/Heavy model routing, an API layer, disabled web-search foundation, and documented security boundaries.

A streaming chat web UI for an agentic RAG assistant — renders responses token-by-token over SSE, backed by the agentic-rag server.

Stateful AI study coach on Cloudflare Agents with persistent plan memory, Workers AI inference, tool calling, streaming chat, and scheduled workflow execution.

A modern full-stack AI chat app built with Next.js and Express, featuring streaming responses, markdown rendering, syntax-highlighted code blocks, and model/provider selection.

Terminal-first TUI for Ollama local LLMs, browse models, pull new ones, chat with system prompts, compare and benchmark, all keyboard-driven.

【Star-crossed coders unite!⭐️】🤖 Reusable template for AI-powered event Q&A assistants. Built with Next.js 16 & Google Gemini. Features full-screen chat, community-driven FAQ voting, Notion backend. Customize in minutes - just edit config files. One-click Vercel deployment.