> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cdp.coinbase.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Hosting Agents

There are several options for hosting your AgentKit agents, from managed AI agent platforms to traditional cloud hosting. This guide covers the main approaches and helps you choose the right one for your needs.

<Tabs groupId="hosting-options">
  <Tab value="autonome" title="Autonome" default>
    [Autonome](https://apps.autono.meme/) by AltLayer provides a managed platform specifically designed for hosting crypto AI agents. It offers:

    * Simple deployment process
    * Built-in monitoring and management
    * Team collaboration features
    * Support for multiple agent frameworks including AgentKit

    #### Quick Setup

    1. Visit [apps.autono.meme](https://apps.autono.meme/)
    2. Log in and create an organization
    3. Click "+" to deploy a new agent
    4. Select "AgentKit" as your framework
    5. Enter required API keys (OpenAI, CDP, etc.)
    6. Deploy your agent

    For detailed instructions, see the [Autonome deployment guide](https://docs.altlayer.io/altlayer-documentation/autonome/deploy-ai-agent).

    #### Custom Framework Upload

    If you have a custom AgentKit implementation, you can upload it as a new framework:

    1. Package your agent as a Docker image
    2. Click "Upload New Framework" in Autonome
    3. Configure the framework details including:
       * Framework name
       * Description
       * Docker image URL
       * Chat endpoint and port
       * Environment variables

    See the [framework upload guide](https://docs.altlayer.io/altlayer-documentation/autonome/uploading-your-own-agent-framework) for complete details.
  </Tab>

  <Tab value="near" title="NEAR AI">
    [NEAR AI](https://near.ai) provides a platform for building and hosting AI agents with built-in support for various frameworks including AgentKit.

    #### Requirements

    * Python 3.9 - 3.11 (3.12 - 3.13 is not supported)
    * NEAR Account (can be created at [wallet.near.org](https://wallet.near.org))
    * Virtual environment (recommended)

    #### Quick Setup

    1. Create and activate a Python virtual environment:

    ```bash lines wrap theme={null}
    # Using venv
    python -m venv .venv
    source .venv/activate

    # Or using conda
    conda create -n nearai python=3.11
    conda activate nearai
    ```

    2. Install the NEAR AI CLI:

    ```bash lines wrap theme={null}
    pip install nearai
    ```

    3. Login to NEAR AI:

    ```bash lines wrap theme={null}
    nearai login
    ```

    4. Create a new agent:

    ```bash lines wrap theme={null}
    nearai agent create
    ```

    5. Configure your agent's framework in `metadata.json`:

    ```json lines wrap theme={null}
    {
      "details": {
        "agent": {
          "framework": "agentkit"
        }
      }
    }
    ```

    For detailed setup instructions, see:

    * [NEAR AI CLI Guide](https://docs.near.ai/cli/)
    * [NEAR AI Quickstart Guide](https://docs.near.ai/agents/quickstart/)
    * [Framework Support Documentation](https://docs.near.ai/agents/env/frameworks/)
  </Tab>

  <Tab value="cloud" title="Cloud Hosting">
    You can host your AgentKit agent on any major cloud platform. Here's a general approach:

    #### AWS Setup

    1. Create an ECS cluster or EKS cluster
    2. Build and push your agent's Docker image to ECR
    3. Deploy using ECS tasks or Kubernetes deployments

    #### Google Cloud Setup

    1. Create a GKE cluster or Cloud Run service
    2. Push your Docker image to Container Registry
    3. Deploy using Cloud Run or Kubernetes
  </Tab>
</Tabs>

## Choosing a Hosting Solution

Consider these factors when selecting a hosting solution:

* **Managed Platforms** (Autonome, NEAR AI): Best for quick deployment and managed infrastructure
* **Cloud Hosting**: Ideal for custom requirements and full control over infrastructure
* **Local Hosting**: Perfect for development and testing

For most users, managed platforms like Autonome or NEAR AI provide the easiest path to production. For enterprise needs or specific requirements, traditional cloud hosting offers more control and customization options.