Large language models have powerful language capabilities but also limitations, such as difficulty handling private data, accessing the latest information, following fixed processes, and automatically planning complex projects. To overcome these limitations, Model Studio provides three types of applications: agent, workflow, and agent orchestration. These applications add features such as knowledge base retrieval, workflow design, and agent collaboration to significantly expand the capabilities of models.
Only users who created Model Studio applications before April 21, 2025 can access the Application tab (as shown in the following figure). This tab contains the following features: applications (agent, workflow, and agent orchestration), components (prompt and plug-in), and data (knowledge base and application data). This feature is in preview. Use with caution in production environments.
RAM user: To perform write operations on the Application tab, such as creating, editing, or publishing applications, or creating a knowledge base, you must obtain the
administratorpage permission.
Only users who created Model Studio applications before April 21, 2025 can call the APIs for features such as application data, knowledge bases, and prompt engineering.
RAM user: To call the APIs for features such as application data, knowledge bases, and prompt engineering, you must obtain the permissions for the corresponding APIs.
Select application type
Agent application (Assistant)
If you are new to model applications, we recommend that you start with the agent application. An agent application is a conversational AI application that completes complex tasks through contextual dialogue, autonomous decision-making, and tool calling. Its simple configuration process and intuitive conversational output format make it ideal for beginners to quickly create single-function AI applications. For more information, see Agent application.
Workflow application (Workflow)
If you are familiar with agent applications but find they do not meet your needs for designing a fixed AI workflow, we recommend using a workflow application. A workflow application is a procedural AI application that breaks down complex tasks into several subtasks to provide greater control over the workflow. You can create custom task flows by dragging and dropping nodes. For more information, see Workflow application.
Feature comparison
Comparison item | Agent application | Workflow application |
Creation method | No-code | No-code |
Application form | Conversational | Procedural |
Dependency on prompts | High | Low |
Supported | Supported | |
Supported (official and custom) | Supported (plug-in node, API node) | |
Text conversion | Partially supported (dependent on prompts) | Supported (text conversion node and script conversion node) |
Integration with external services | Supported (custom plugins) | Supported (Function Compute node) |
Branching | Partially supported (dependent on prompts) | Supported (logical branching and intent recognition) |
Automatic planning | Partially supported (dependent on prompts) | Supported (agent group node) |