| Github | Huggingface | Pypi | Open Source AI Agent Marketplace DeepNLP | Agent RL Dataset |
This is the official github repo for pypi package ai_agent_marketplace https://pypi.org/project/ai-agent-marketplace. The repo Open Source AI Agent Marketplace and deepnlp.org website AI Agent Marketplace Store & Search Engine aims to provides a public repo and index of more than 10K+ AI Agent information from 30+ categories in the communities, such as autonomous agent, chatbots, Computer and Mobile phone use agents, robotic agents, and various industries such as business, finance, law, medical or healthcare, etc. The directory are updated to websites from both public repo (github/huggingface) as well as AI Agent services in cloud service provider (Microsoft Azure AWS, Copilot, OpenAI Agent app Store GPT Store, Google Cloud, etc).
AI Agent Marketplace Distribution By Category

We would like to help developers and end users within the lifecycle of AI agent development. From the registration, deployment, Agent router, API calling, metric/traffic tracking, and finally to the stage of moneyterization from your AI Agent. Anyone can submit their AI agents card information, code, APIs, pricing plans to the public registry, just like your submit a paper to arxiv.org and submit models to huggingface.co.
graph LR
Dev[Developers]--> A[AI Agent/MCP Development] --> B[Deloyment] --> C[AI Agent Registry to Marketplace] --> D[Agent Search and Exploration]
U[Users] --> R[Agent Router & Data Tracking] --> F[API Billing System] --> G[Feedback Comments & Discussion] --> H[Moneytization]
AI Agent Marketplace and Search Engine

AI Agent Registration

AI Agent Calls Credit Usage

Traffic Tracking

OneKey Router

Users Genunie Reviews

List of Methods that you can register your AI agent
| method | usage |
|---|---|
| Website | Visit the official AI Agent registry |
| curl | Support to Submit your Github Repo contents to the marketplace |
| CLI | Command Line agtm upload --github or agtm upload --config ./agent.json |
| python | Install packages pip install ai-agent-marketplace |
| nodejs | Install packages npm install -g @aiagenta2z/agtm |
Best suitable when you already put your AI Agent introduction on github and you can just curl to the registry.
Let’s say you want to submit an MCP github repo, we use the markitdown repo for example: https://github.com/microsoft/markitdown.
Firstly, get the keys from Keys Generation generate AI_AGENT_MARKETPLACE_ACCESS_KEY as developer.
curl -X POST https://www.deepnlp.org/api/ai_agent_marketplace/registry -H "Content-Type: application/json" -d '{"github":"https://github.com/microsoft/markitdown", "access_key":"{AI_AGENT_MARKETPLACE_ACCESS_KEY}"}'
Install the command line using pip or nodejs first, Get access key at keys
export AI_AGENT_MARKETPLACE_ACCESS_KEY="{your_access_key}"
agtm upload --github https://github.com/AI-Hub-Admin/My-First-AI-Coding-Agent
## upload from json file or yaml file
agtm upload --config ./agent.json
agtm upload --config ./agent.yaml
Demo examples can be found in ./agent.json or ./agent.yaml
Setup Your Own Endpoint or Schema
Please visit the command line github package agtm detailed usage
agtm upload --config ./agent.json --endpoint https://www.example.com --schema ./schema.json
For test API Key, please set variable of AI_AGENT_MARKETPLACE_ACCESS_KEY
export AI_AGENT_MARKETPLACE_ACCESS_KEY="TEST_KEY_AI_AGENT_REGISTRY"
agtm upload --config ./agent.json --endpoint https://www.deepnlp.org/api/ai_agent_marketplace/registry --schema ./schema.json
# or
agtm upload --config ./agent.json --endpoint https://www.aiagenta2z.com/api/ai_agent_marketplace/registry --schema ./schema.json
pip install ai-agent-marketplace
Get Keys and Register your AI Agent
export AI_AGENT_MARKETPLACE_API_KEY={Your API Key}
submit your AI Agent
import ai_agent_marketplace as aa
import json
def publish_your_agent():
"""
access_key can be obtained from your personal page:
www.deepnlp.orgworkspace/my_ai_services
once you submit, it's pending approval and you can track the data then
get your access_key from http://www.deepnlp.org/workspace/my_ai_services
"""
access_key = "${your_access_key}"
name = "My First AI Agent"
item_info = {}
item_info["content"] = "This AI Agent can do complicated programming work for humans"
item_info["website"] = "https://www.my_first_agent.com"
item_info["field"] = "AI AGENT"
item_info["subfield"] = "Coding Agent"
item_info["content_tag_list"] = "coding,python"
result = aa.add(access_key=access_key, name="My First Agent", item_info=item_info)
url = result["url"] if "url" in result else ""
msg = result["msg"] if "msg" in result else ""
print ("## DEBUG: AI Agent Marketplace Post msg is|%s" % str(msg))
print ("## DEBUG: AI Agent Marketplace Post url is|%s" % str(url))
publish_your_agent()
We provides both API and python package wrapper
import ai_agent_marketplace as aa
import json
def search_ai_agent_traffic_data():
result = aa.search(q="Coding Agent Jetbrains")
print ("## DEBUG: search result is|%s" % str(result))
result2 = aa.search(q="Coding Agent", limit=20, timeout=5)
print ("## DEBUG: search result is|%s" % str(result2))
result3 = aa.search(q="", limit=20, timeout=5)
print ("## DEBUG: search result is|%s" % str(result3))
search_ai_agent_traffic_data()
For example, we use the AI Agent MCP server Google Maps MCP Servers as example.
The unique_id should follow the same /{owerid}/{item-id} format
{
"unique_id": "google-maps/google-maps",
"content_name": "Google Maps MCPs",
"content": "Google Maps MCPs provides Location Service to support various APIs..."
"category": "Map",
"field": "MCP SERVER",
"subfield": "Map",
"website": "maps.google.com",
"content_tag_list": "official,maps,location",
"github": "https://github.com/modelcontextprotocol/servers/tree/main/src/google-maps"
}
See the document for node CLI wrapper at NPM Agtm package
Try Web App of Onekey API Router Agent
Some of the key metric that DeepNLP AI Agent Marketplace monitors include:

You can visit DeepNLP AI Agent Marketplace and Directory to find more.
Suppose you want to find out some suitable AI Employees which can help you write reports or emails, schedule meeting, write codes, etc. You may use Google or Bing to search keywords like “AI Agent Employees”, “Report Writing Agents”, “Coding Agents”, etc. And DeepNLP AI Agents Marketplace helps you monitor the AI Agents search results ranking. You can navigate to the “Email writing” tab of the AI Agent Directory and see the top reviewed AI Agents and its latested Bing Search Ranking.
And you can navigate the Email Writing AI Agents Directory and find out some top rated AI Agents in Email Writing such as

mailmeteor.com Bing Search Rank 2.0
galaxy.ai Bing Search Rank 3.0
The metric of Bing search results of AI agents are calculated on various related keywords and updated daily. And you can monitor the daily trends of fast growing AI agents.

Some of the AI Agents are public available and you can use github to help you track the performance and help you find the best open source AI agents. You can visit the AI Agent Directory of Coding Agents and help you monitor the performance such as github repo stars. List of AI Coding Agent

Can you can find the popular stared Github projects such as:


| Snapshot | Dataset |
| 2025-10 | https://huggingface.co/datasets/DeepNLP/AI-Agent-Marketplace-Index |
Please contribute to the AGENT.md to include links and introduction to your repo.