Agents and tools that carry out tasks with limited human input, from general-purpose AI agents and browser automation to open-source frameworks and no-code workflow platforms.
Agents that plan and carry out multi-step tasks with limited supervision. This section covers hosted general-purpose agents, agents that operate a browser or desktop directly, and open-source frameworks for building your own.
Hosted agents that take a goal and autonomously plan, use tools, and return a result.
| Tool | Description | License | Pricing |
|---|---|---|---|
| ChatGPT Agent | OpenAI's agent mode in ChatGPT, which can browse, run code, and use tools to complete tasks end to end. | ||
| Genspark | A general-purpose agent for research, document, and media tasks, with its own built-in tool set. | ||
| Hermes Agent | An open-source, self-hosted agent from Nous Research with persistent memory that improves as it runs. | ||
| Manus | A general-purpose agent that plans and executes multi-step tasks such as research, analysis, and content production. | ||
| Microsoft Copilot | Microsoft's assistant with an agent mode that completes tasks across Microsoft 365 apps. | ||
| OpenClaw | An open-source, self-hosted agent that connects LLMs to your apps, browser, and system tools, controlled from a chat app. |
Agents that act directly in a web browser or on the desktop, clicking, typing, and navigating applications on your behalf.
| Tool | Description | License | Pricing |
|---|---|---|---|
| OpenAI Atlas | OpenAI's browser with an agent mode that completes multi-step tasks on the web. | ||
| Perplexity Comet | Perplexity's browser with an assistant that researches, summarizes, and completes tasks within the page. | ||
| Claude (Computer Use) | Anthropic's Claude controlling a browser or desktop to complete tasks, via Claude in Chrome and Claude Cowork. | ||
| Google Project Mariner | Google DeepMind's agent that navigates and acts within the browser, available in the Gemini app. | ||
| Browser Use | An open-source framework that lets AI models control a browser for custom automation. |
Open-source libraries for building, orchestrating, and running your own agents.
| Tool | Description | License | Pricing |
|---|---|---|---|
| AutoGPT | One of the first autonomous-agent projects, now a platform for building and running agents. | ||
| CrewAI | A framework for orchestrating role-based teams of collaborating agents. | ||
| LangGraph | A framework from the LangChain ecosystem for building stateful, controllable agents as graphs. | ||
| Microsoft AutoGen | A Microsoft framework for multi-agent conversational systems. | ||
| OpenAI Agents SDK | OpenAI's lightweight SDK for building agents with tools, handoffs, and guardrails. | ||
| smolagents | A minimal agent library from Hugging Face, focused on code-writing agents. | ||
| Google ADK | Google's Agent Development Kit for building and deploying agents. |
Platforms for connecting apps and automating workflows, increasingly with AI steps built in.
| Tool | Description | License | Pricing |
|---|---|---|---|
| Activepieces | An open-source workflow automation platform with a visual builder. | ||
| Bardeen | A no-code automation tool for repetitive tasks across web applications. | ||
| Gumloop | An AI-native platform for building automations by connecting modular nodes on a canvas. | ||
| Lindy | A no-code platform for building AI assistants that automate business workflows. | ||
| Make | A visual platform for building multi-step automations across thousands of apps. | ||
| n8n | An open-source, self-hostable workflow automation tool with a visual editor. | ||
| Relay.app | An automation platform that combines workflows with AI steps and human approvals. | ||
| Zapier | A widely used automation platform connecting thousands of apps with no-code "Zaps". |
