# Overview

## Overview

The AI landscape has exploded with powerful platforms that go beyond simple model inference. From autonomous agents that write code and solve problems, to visual workflow builders that let you create AI applications without coding, to full-featured chat interfaces that rival commercial offerings — all of these can be self-hosted on Clore.ai's GPU infrastructure.

This section covers **three categories** of AI platforms:

### 🤖 AI Agent Frameworks

Autonomous agents that can plan, execute tasks, write code, and interact with tools:

| Platform                                                                    | Stars | GPU Required   | Description                                                |
| --------------------------------------------------------------------------- | ----- | -------------- | ---------------------------------------------------------- |
| [AutoGPT](https://docs.clore.ai/guides/ai-platforms-and-agents/autogpt)     | 175K+ | No (API-based) | The original autonomous AI agent platform                  |
| [CrewAI](https://docs.clore.ai/guides/ai-platforms-and-agents/crewai)       | 44K+  | No (API-based) | Multi-agent orchestration with role-playing agents         |
| [MetaGPT](https://docs.clore.ai/guides/ai-platforms-and-agents/metagpt)     | 45K+  | No (API-based) | Simulates a software company with specialized agent roles  |
| [SuperAGI](https://docs.clore.ai/guides/ai-platforms-and-agents/superagi)   | 15K+  | Optional       | Dev-first autonomous agent framework with GUI              |
| [OpenHands](https://docs.clore.ai/guides/ai-platforms-and-agents/openhands) | 65K+  | No (API-based) | AI-powered software development agent (formerly OpenDevin) |
| [SWE-agent](https://docs.clore.ai/guides/ai-platforms-and-agents/swe-agent) | 15K+  | No (API-based) | Automated GitHub issue fixing agent                        |

### 🔧 AI Workflow Builders

Visual and low-code platforms for building AI applications and automations:

| Platform                                                                  | Stars | GPU Required | Description                                               |
| ------------------------------------------------------------------------- | ----- | ------------ | --------------------------------------------------------- |
| [Dify.ai](https://docs.clore.ai/guides/ai-platforms-and-agents/dify)      | 114K+ | Optional     | Production-ready AI workflow and agent platform           |
| [n8n](https://docs.clore.ai/guides/ai-platforms-and-agents/n8n)           | 55K+  | No           | Fair-code workflow automation with native AI capabilities |
| [Langflow](https://docs.clore.ai/guides/ai-platforms-and-agents/langflow) | 55K+  | No           | Visual drag-and-drop AI application builder               |
| [Flowise](https://docs.clore.ai/guides/ai-platforms-and-agents/flowise)   | 35K+  | No           | Visual AI agent and chatbot builder                       |
| [Haystack](https://docs.clore.ai/guides/ai-platforms-and-agents/haystack) | 18K+  | Optional     | AI orchestration framework for RAG and search             |

### 💬 Self-Hosted AI Assistants

Full-featured chat interfaces and AI assistant platforms:

| Platform                                                                        | Stars | GPU Required   | Description                                          |
| ------------------------------------------------------------------------------- | ----- | -------------- | ---------------------------------------------------- |
| [LobeChat](https://docs.clore.ai/guides/ai-platforms-and-agents/lobechat)       | 55K+  | No (API-based) | Modern AI chat framework with multi-provider support |
| [LibreChat](https://docs.clore.ai/guides/ai-platforms-and-agents/librechat)     | 22K+  | No (API-based) | Enhanced ChatGPT clone with multiple AI providers    |
| [AnythingLLM](https://docs.clore.ai/guides/ai-platforms-and-agents/anythingllm) | 40K+  | No             | All-in-one AI app with RAG and agent builder         |
| [Jan.ai](https://docs.clore.ai/guides/ai-platforms-and-agents/jan)              | 40K+  | Yes            | Offline-first AI assistant with local model support  |
| [GPT4All](https://docs.clore.ai/guides/ai-platforms-and-agents/gpt4all)         | 72K+  | Optional       | Privacy-focused local LLM runner                     |

### 💻 AI Coding Tools

Tools for AI-powered code completion, review, and development:

| Platform                                                                          | Stars | GPU Required  | Description                              |
| --------------------------------------------------------------------------------- | ----- | ------------- | ---------------------------------------- |
| [Continue.dev](https://docs.clore.ai/guides/ai-platforms-and-agents/continue-dev) | 25K+  | Yes (backend) | Open-source AI coding assistant for IDEs |

> **Already covered in other sections:** [Open WebUI](https://docs.clore.ai/guides/language-models/open-webui), [LocalAI](https://docs.clore.ai/guides/language-models/localai-openai-compatible), [Text Generation WebUI](https://docs.clore.ai/guides/language-models/text-generation-webui), [TabbyML](https://docs.clore.ai/guides/ai-coding-tools/tabby), [Aider](https://docs.clore.ai/guides/ai-coding-tools/aider)

## Why Run AI Platforms on Clore.ai?

### 💰 Cost Efficiency

Most AI agent platforms themselves are lightweight (CPU-only), but they become powerful when connected to local GPU-accelerated LLMs. On Clore.ai, you can rent an RTX 4090 for \~$0.35/hr — a fraction of cloud API costs for heavy workloads.

### 🐳 Docker Native

Every platform in this section supports Docker deployment. Clore.ai servers come with Docker and NVIDIA drivers pre-installed, so you can be up and running in minutes.

### 🔒 Data Privacy

Self-hosting means your data, prompts, and conversations never leave your rented server. Perfect for enterprises and privacy-conscious users.

### ⚡ Flexible Architecture

Run the AI platform and the LLM on the same server, or split them across machines. Use API-based models for quick experiments, then switch to local models for production.

## Common Patterns

### Pattern 1: AI Platform + Cloud APIs

Rent a cheap CPU server on Clore.ai, deploy your platform (Dify, n8n, etc.), and connect to OpenAI/Anthropic APIs. Lowest cost for light usage.

### Pattern 2: AI Platform + Local LLM (Same Server)

Rent a GPU server, run both the platform and Ollama/vLLM on the same machine. Best for privacy and consistent performance.

### Pattern 3: AI Platform + Dedicated LLM Server

Run the platform on one server and the LLM on a separate GPU server. Best for scaling and team usage.

## Quick Reference

| If you want to...                   | Use this                                                                                                                                                                                                              |
| ----------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Build AI workflows visually         | [Dify](https://docs.clore.ai/guides/ai-platforms-and-agents/dify), [Langflow](https://docs.clore.ai/guides/ai-platforms-and-agents/langflow), [Flowise](https://docs.clore.ai/guides/ai-platforms-and-agents/flowise) |
| Automate business processes with AI | [n8n](https://docs.clore.ai/guides/ai-platforms-and-agents/n8n)                                                                                                                                                       |
| Run autonomous coding agents        | [OpenHands](https://docs.clore.ai/guides/ai-platforms-and-agents/openhands), [SWE-agent](https://docs.clore.ai/guides/ai-platforms-and-agents/swe-agent)                                                              |
| Create multi-agent teams            | [CrewAI](https://docs.clore.ai/guides/ai-platforms-and-agents/crewai), [MetaGPT](https://docs.clore.ai/guides/ai-platforms-and-agents/metagpt)                                                                        |
| Self-host a ChatGPT alternative     | [LobeChat](https://docs.clore.ai/guides/ai-platforms-and-agents/lobechat), [LibreChat](https://docs.clore.ai/guides/ai-platforms-and-agents/librechat)                                                                |
| Chat with your documents (RAG)      | [AnythingLLM](https://docs.clore.ai/guides/ai-platforms-and-agents/anythingllm), [Dify](https://docs.clore.ai/guides/ai-platforms-and-agents/dify)                                                                    |
| AI code completion in your IDE      | [Continue.dev](https://docs.clore.ai/guides/ai-platforms-and-agents/continue-dev), [TabbyML](https://docs.clore.ai/guides/ai-coding-tools/tabby)                                                                      |
| Run models 100% offline             | [Jan.ai](https://docs.clore.ai/guides/ai-platforms-and-agents/jan), [GPT4All](https://docs.clore.ai/guides/ai-platforms-and-agents/gpt4all)                                                                           |

## Getting Started

1. **Choose your platform** from the guides above
2. **Rent a server** on [Clore.ai](https://clore.ai) — see [GPU Comparison](https://docs.clore.ai/guides/getting-started/gpu-comparison) for help choosing
3. **Follow the guide** — most platforms can be deployed with a single `docker compose up`
4. **Connect models** — use [Ollama](https://docs.clore.ai/guides/language-models/ollama) or [vLLM](https://docs.clore.ai/guides/language-models/vllm) for local inference, or plug in your API keys

> **Tip:** Start with [Dify](https://docs.clore.ai/guides/ai-platforms-and-agents/dify) or [n8n](https://docs.clore.ai/guides/ai-platforms-and-agents/n8n) if you're new to AI platforms — they offer the best balance of power and ease of use.