Official plugins

First-time access to the plugins page

If your Alibaba Cloud account or RAM user (sub-account) does not have the AliyunServiceRoleForSFMAccessCloudAPI role, you cannot access plugins. Follow these instructions to grant the required permissions.

Alibaba Cloud account
RAM user (sub-account)

If you log in to Model Studio with an Alibaba Cloud account, go to the Plugins page, select the checkbox to agree to the terms, and click Authorize and Enter.The authorization page creates the service-linked role AliyunServiceRoleForSFMAccessCloudAPI, which is associated with the AliyunServiceRolePolicyForSFMAccessCloudAPI policy. This role authorizes Model Studio to access your Alibaba Cloud Marketplace products and make API calls based on the plugin configuration.

If you log in to Model Studio with a RAM user (sub-account), go to the Plugins page, select the checkbox to agree to the terms, and click Authorize and Enter. The following prompt appears:A dialog box for the Cloud Marketplace service-linked role appears, displaying error code 140052 and the service-linked role name AliyunServiceRoleForSFMAccessCloudAPI.This error occurs because the RAM user does not have permission to create a service-linked role. Follow the steps below to grant this permission to the RAM user. After you grant this permission, the RAM user can access the Plug-in page.

Grant the RAM user the permission to create a service-linked role.
1. Log in to the RAM console with your Alibaba Cloud account.
2. In the left-side navigation pane, choose Permissions>Policy.
3. Click Create Policy.
4. On the Script Editor tab, enter the following content in the Effect, Action, Resource, and Condition fields.

{
    "Action": [
        "ram:CreateServiceLinkedRole"
    ],
    "Resource": "*",
    "Effect": "Allow",
    "Condition": {
        "StringEquals": {
            "ram:ServiceName": "cloundapi-access.sfm.aliyuncs.com"
        }
    }
}

Click OK.
Set a name for the policy and click OK. For example, enter the name service-linked role.
In the left-side navigation pane, choose Identities > Users.
Find the RAM user to authorize and, in the Actions column, click Add Permission.
Select the policy that you created and click OK. The RAM user can now create service-linked roles. From the policy type drop-down list, select Custom Policy to filter the list. Then, find and select your target policy.
Return to the Plugins page, select the checkbox to agree to the terms, and click Authorize and Enter.

The Component Market provides pre-built official plugins from Model Studio. You can call them directly without configuring their input and output parameters.

Plugin name	Tool ID	Description	Billing
Python Code Interpreter	code_interpreter	Enables the model to run Python code for tasks like mathematical calculations, data analysis, visualization, and data processing.	Free
Calculator	calculator	Lets the model perform complex mathematical calculations, such as "12313 * 13232".	Free
Image Generation	text_to_image	Lets the model generate images from text, such as "Draw a smiling dog."	Free for a limited time. Activation is required.

Python Code Interpreter

Sample input Output without plugin Output with plugin

Sample input	Output without plugin	Output with plugin
`import matplotlib.pyplot as plt import numpy as np x = np.linspace(-2, 2, 100) y1 = np.sqrt(1 + x2) y2 = -np.sqrt(1 + x2) plt.plot(x, y1, label='y = sqrt(1 + x^2)') plt.plot(x, y2, label='y = -sqrt(1 + x^2)') plt.xlabel('x') plt.ylabel('y') plt.title('Hyperbola') plt.legend() plt.show()`	The agent application cannot execute Python code; it can only provide a text description of the code.	The agent application can execute Python code for data visualization and analysis.

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(-2, 2, 100)
y1 = np.sqrt(1 + x**2)
y2 = -np.sqrt(1 + x**2)

plt.plot(x, y1, label='y = sqrt(1 + x^2)')
plt.plot(x, y2, label='y = -sqrt(1 + x^2)')
plt.xlabel('x')
plt.ylabel('y')
plt.title('Hyperbola')
plt.legend()
plt.show()

The agent application cannot execute Python code; it can only provide a text description of the code.

The agent application can execute Python code for data visualization and analysis.

The Python Code Interpreter plugin does not support accessing external networks or uploading local files. The available dependencies are: json5~=0.9.6, jupyter_client~=8.1.0, ipykernel~=6.25.0, seaborn, sympy, pydantic~=1.10.8, pillow~=9.4.0, fastapi~=0.101.1, dynaconf~=3.2.1, oss2~=2.18.1, matplotlib, starlette~=0.27.0, uvicorn~=0.23.2, requests~=2.31.0, scipy, html2text, pandas, pdf2image, pdfminer-six, pypdf, python-pptx, and wordcloud.

Calculator

Sample input	Output without plugin	Output with plugin
12313*13232	The agent application cannot reliably solve complex math problems and may return incorrect results. The correct answer is 162,925,616.	The agent application has strong calculation capabilities and can generate accurate results.

Image Generation

Sample input	Output without plugin	Output with plugin
Draw a smiling dog.	The agent application cannot draw images or graphics directly and can only describe the scene in text.	The agent application can draw images directly and can generate a picture of a dog based on the instructions.
Draw a smiling dog.

Combine plugins

Model Studio allows you to call multiple tools in a single task. You can select plugins to suit your needs.

Call official plugins

Step 1. Authorize the workspace (if required). You must authorize an official plugin before you can use it in a sub-workspace. This step is not required for the default workspace.

Go to the Plugins page, find the target plugin, and click View Details.
Click Authorization, select the sub-workspace to authorize, and click OK to complete the authorization.

Step 2.Call a plugin.

Call a plugin

Method 1: From the Plug-in page, add a tool to an agent application.
An official plugin can only be added to an Agent Application in the same workspace.
1. Find the target plugin and click Add to Agent.
2. Select the tool and click Next.
3. Select an agent application and click Confirm Add.
4. On the application details page, the tool now appears automatically.
  You can also click Select Plug-in to add more tools. You can add up to 10 tools. The agent application selects and calls one or more tools based on the input.
5. Test the tool by interacting with the model in the input box.
6. After testing, publish the application.
Method 2: Go to the Apps page. In a specific agent or workflow application, add the plugin, test its functionality, and then Publish the application. For more information, see Agent application plugin capabilities and Workflow application plugin node.

Get the tool ID

A tool ID identifies a specific tool. You must provide the correct tool ID when calling a tool by using an API.

On the Plug-in page, find the target plugin and click View Details.
Under Plug-in Tool, get the tool ID. In this example, the tool ID is calculator.

In addition to official plugins, Model Studio supports custom plugins. For more information, see Custom plugins.

​First-time access to the plugins page