Awesome AI Agents › Low-Code/No-Code Platforms
LMForge, published as OpenAgent, is an end-to-end platform for building, orchestrating, publishing and operating LLM applications rather than a single chat demo. The backend is Python with Flask, SQLAlchemy, Celery workers, Flask-SocketIO, Redis and PostgreSQL, and the agent layer is built on LangChain and LangGraph. Vector search uses Weaviate and FAISS with lexical, semantic and hybrid retrieval modes. The frontend is a Vue 3 workspace built with Vite, Pinia, TailwindCSS, Vue Flow and Arco Design. Users build AI applications from a dedicated workspace that supports drafts, publishing, usage analysis, version comparison and prompt comparison. A visual workflow editor provides nodes for LLM calls, tool invocation, dataset retrieval, code execution, HTTP requests, conditional branching, text processing, template transforms and structured parameter extraction. A Deep Research mode lets an application decompose a complex request and coordinate its bound capabilities across multiple execution steps. A home assistant entry point routes a natural-language request either to an already published public agent through the A2A protocol or into a flow that creates a new application from the stated requirement. Dataset management covers document upload, segment inspection and binding retrieval to agents and workflows. Store-style views let users browse public applications, tools and workflows contributed by others. Published applications are exposed over REST and server-sent events through an OpenAPI chat endpoint, so external systems can call them directly. Deployment is via Docker Compose with Nginx in front, and the stack is documented as needing roughly 8 GB of RAM. Model support spans OpenAI, DeepSeek, Grok, Google, Moonshot, Tongyi, Wenxin, Zhipu, Atlas Cloud and local Ollama endpoints. The project targets teams that want a self-hosted alternative to hosted agent-building platforms, with both Chinese and English documentation. Licensed MIT.
https://github.com/Haohao-end/LMForge-End-to-End-LLMOps-Platform-for-Multi-Model-Agents
AnythingLLM is an all-in-one AI application that enables intelligent chat interactions with documents using various large language models, supporting multi-user management, no-code AI agent building, and multi-modal capabilities in both desktop and Docker environments.
NocoBase is an open-source AI + no-code platform that accelerates business system development by combining AI agents with a WYSIWYG no-code interface for speed and reliability.
Maxun is an open-source no-code platform that enables users to create custom robots for automated web data extraction, turning websites into APIs and spreadsheets without coding.
Databerry is a no-code platform that enables users to build custom Large Language Model agents without programming skills, making AI development accessible and user-friendly.
LLMStack is a no-code multi-agent framework for building generative AI agents, workflows, and applications that integrate with your data and business processes, deployable on cloud or on-premise with multi-tenant support and API access.
Source-available agent platform that pairs a drag-and-drop visual builder with a TypeScript SDK kept in two-way sync, plus a configuration API, CLI, chat UI library and OpenTelemetry tracing.
Next.js template for building an AI workflow automation platform, with a React Flow drag-and-drop canvas, prebuilt integrations, execution logs and conversion of workflows into TypeScript code.
Heym is a source-available, self-hosted workflow automation platform where LLMs and agents are the execution model, combining a drag-and-drop canvas with RAG, MCP, human-in-the-loop steps, evals and cost tracking.