Ai Agent

Mem0 is a memory layer for AI applications and agents. It focuses on extracting, storing and retrieving useful user or interaction context so applications can become more personalized across sessions without placing all history directly into every prompt. It can complement agent frameworks rather than replace them, and is most valuable when an application has a clear policy for what should be remembered, updated or forgotten.

Mem0 is a memory layer for AI applications and agents. It focuses on extracting, storing and retrieving useful user or interaction context so applications can become more personalized across sessions without placing all history directly into every prompt. It can complement agent frameworks rather than replace them, and is most valuable when an application has a clear policy for what should be remembered, updated or forgotten.

CrewAI is a Python framework for orchestrating collaborative AI agents, with an emphasis on assigning roles, tasks and coordinated workflows to groups of agents. It is useful when a product needs explicit multi-agent decomposition and reusable orchestration rather than only a single tool-calling loop. The framework has a large ecosystem and active adoption, but successful production use still requires careful workflow design, observability and limits on autonomous behavior.

CrewAI is a Python framework for orchestrating collaborative AI agents, with an emphasis on assigning roles, tasks and coordinated workflows to groups of agents. It is useful when a product needs explicit multi-agent decomposition and reusable orchestration rather than only a single tool-calling loop. The framework has a large ecosystem and active adoption, but successful production use still requires careful workflow design, observability and limits on autonomous behavior.

AI Agent Assistant & development framework that integrates lots of IM platforms, LLMs, plugins and AI feature, and can be your openclaw alternative. ✨

Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.

Low-code framework for building custom LLMs, neural networks, and other AI models

Hermes Agent is a personal AI agent platform built around persistent memory, skills learned from experience, tool use and access through multiple interfaces. The project supports a CLI/TUI, messaging gateways, scheduled tasks, MCP integration, browser and terminal tools, multiple model providers and self-hosted operation. Its differentiator is the long-lived personal-agent model: retaining useful context and procedural skills across sessions rather than treating every run as a stateless request.

Hermes Agent is a personal AI agent platform built around persistent memory, skills learned from experience, tool use and access through multiple interfaces. The project supports a CLI/TUI, messaging gateways, scheduled tasks, MCP integration, browser and terminal tools, multiple model providers and self-hosted operation. Its differentiator is the long-lived personal-agent model: retaining useful context and procedural skills across sessions rather than treating every run as a stateless request.

Vercel AI SDK is a provider-agnostic TypeScript toolkit for building AI applications and agents across popular JavaScript frameworks and runtimes. Its strongest fit is application-layer integration: unified model calls, tool use, streaming, structured output, chat and generative UI primitives, plus adapters for multiple model providers. It is best suited to developers who want one application API across model vendors rather than a full autonomous-agent runtime or infrastructure platform.

Vercel AI SDK is a provider-agnostic TypeScript toolkit for building AI applications and agents across popular JavaScript frameworks and runtimes. Its strongest fit is application-layer integration: unified model calls, tool use, streaming, structured output, chat and generative UI primitives, plus adapters for multiple model providers. It is best suited to developers who want one application API across model vendors rather than a full autonomous-agent runtime or infrastructure platform.

💖🧸 Self hosted, you-owned Grok Companion, a container of souls of waifu, cyber livings to bring them into our worlds, wishing to achieve Neuro-sama's altitude. Capable of realtime voice chat, Minecraft, Factorio playing. Web / macOS / Windows supported.

Activepieces is an open-source automation platform centered on a visual flow builder and an extensible TypeScript integration system called pieces. It targets both technical and non-technical automation users, supports branching and loops, and has increasingly strong AI and MCP-oriented integrations. It is a good fit when a team wants a self-hostable workflow product with a user-facing builder rather than only a code-first orchestration library.

A lightweight, powerful framework for multi-agent workflows

Agent Zero is a general-purpose autonomous agent platform designed to run as an interactive, tool-using environment rather than a narrow library. Its current documentation emphasizes Docker-based installation, a Web UI, model configuration, host connectivity and practical workflows. It fits users who want an extensible agent workspace with terminal and host integration, while the broad capability surface means deployments need careful permissions and operational boundaries.

Agent Zero is a general-purpose autonomous agent platform designed to run as an interactive, tool-using environment rather than a narrow library. Its current documentation emphasizes Docker-based installation, a Web UI, model configuration, host connectivity and practical workflows. It fits users who want an extensible agent workspace with terminal and host integration, while the broad capability surface means deployments need careful permissions and operational boundaries.

The ultimate LLM/AI application development framework in Go.

Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat apps

TradingAgents: Multi-Agents LLM Financial Trading Framework

AgenticSeek is an autonomous assistant project oriented toward private or locally controlled agent workflows, including research and task execution. It is positioned closer to an end-user agent system than a small SDK, making it interesting for users who want a self-hosted assistant stack with less dependence on a hosted agent product. As with other broad autonomous assistants, actual capability and reliability depend on the configured models, local resources, browser/tool integrations and task boundaries.

AgenticSeek is an autonomous assistant project oriented toward private or locally controlled agent workflows, including research and task execution. It is positioned closer to an end-user agent system than a small SDK, making it interesting for users who want a self-hosted assistant stack with less dependence on a hosted agent product. As with other broad autonomous assistants, actual capability and reliability depend on the configured models, local resources, browser/tool integrations and task boundaries.

Mastra is the modern TypeScript framework for AI-powered applications and agents.

Microsoft AutoGen is a programming framework for agentic AI and multi-agent systems. It provides abstractions for building agents that exchange messages, use tools and participate in structured orchestration patterns, making it suitable for developers who want a flexible research-to-application framework rather than a fixed end-user automation product. Its breadth is valuable for experimentation, but teams should expect to design their own application policies, evaluation and operational controls around it.

Microsoft AutoGen is a programming framework for agentic AI and multi-agent systems. It provides abstractions for building agents that exchange messages, use tools and participate in structured orchestration patterns, making it suitable for developers who want a flexible research-to-application framework rather than a fixed end-user automation product. Its breadth is valuable for experimentation, but teams should expect to design their own application policies, evaluation and operational controls around it.

Pocket Flow: 100-line LLM framework. Let Agents build Agents!

IntentKit is an open-source, self-hosted cloud agent cluster that manages a collaborative team of AI agents for you.

🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming

Open-AutoGLM is a mobile-device agent framework that combines visual understanding, planning and device automation to operate phone interfaces. Its documented Android path uses ADB for control, with model-driven perception of the screen and generated action sequences. It is most relevant to research and controlled automation scenarios where an agent must interact with existing mobile user interfaces rather than only call application APIs.

Open-AutoGLM is a mobile-device agent framework that combines visual understanding, planning and device automation to operate phone interfaces. Its documented Android path uses ADB for control, with model-driven perception of the screen and generated action sequences. It is most relevant to research and controlled automation scenarios where an agent must interact with existing mobile user interfaces rather than only call application APIs.

A self-hosted email client with an AI agent, running entirely on Cloudflare Workers

What are the principles we can use to build LLM-powered software that is actually good enough to put in the hands of production customers?