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

# Cursor

> AI-driven code editor integration guide

Cursor is an AI-driven code editor. By integrating APIYI, you can get powerful AI assistance while writing code.

## Quick Configuration

### 1. Open Settings

Click the gear icon ⚙️ in the upper right corner, select **Models** option

### 2. Configure API

* **OpenAI API Key**: Enter your APIYI key (can use default token directly)
* **Override OpenAI Base URL**: Check and enter `https://api.apiyi.com/v1`
* Click **Verify** to validate configuration

<img src="https://mintcdn.com/apiyillc/OMY6ItCc2mC1yzgA/images/cursor-setting.png?fit=max&auto=format&n=OMY6ItCc2mC1yzgA&q=85&s=47cf7afe1c00e511395e34cc033b877e" alt="Cursor Configuration Interface" width="2066" height="990" data-path="images/cursor-setting.png" />

### 3. Model Configuration

<Warning>
  **Important Note**: Cursor currently **does not support Agent mode**, only Chat conversation mode is available. You can have AI generate code in conversation, then manually apply it to actual code.

  If you're a beginner and heavily rely on Agent mode for Vibe Coding, recommend:

  * Purchase Cursor official membership to use their native service
  * Or use alternatives: VS Code's **RooCode** or **Cline** plugins (support Agent mode)
</Warning>

#### Recommended Model Configuration

Based on current latest model performance and cost-effectiveness, recommend the following:

**Programming First Choice Models**:

* `claude-sonnet-4-20250514` - Claude 4 Sonnet, strongest programming
* `gpt-4.1` - Fast speed, strong overall capability
* `deepseek-v3` - Excellent Chinese programming, high cost-performance

**Cost Optimization Models**:

* `gpt-4.1-mini` - Lightweight but capable
* `claude-3-haiku` - Cheapest in Claude series
* `gemini-2.5-flash` - Google's fast response model

**Reasoning Enhanced Models**:

* `o4-mini` - First choice reasoning model for programming tasks
* `o3` - Complex reasoning and algorithmic problems

#### Add Custom Models

Add the following model IDs in Cursor settings:

```
claude-sonnet-4-20250514
gpt-4.1
deepseek-v3
o4-mini
gemini-2.5-pro
```

## Usage Mode Explanation

### Chat Mode Workflow

Since Cursor doesn't support Agent mode, recommend the following workflow:

1. **Dialogue to Generate Code**
   * Use `Ctrl/Cmd + L` to open chat
   * Describe your requirements, have AI generate code
   * View generated code snippets

2. **Manually Apply Code**
   * Copy code from chat window
   * Paste to target file
   * Or use "Apply" button (if available)

3. **Iterate and Optimize**
   * Continue conversation to request modifications
   * Repeat application process

### Alternative Comparison

| Tool                   | Agent Mode | Advantages                                      | Disadvantages                  |
| ---------------------- | ---------- | ----------------------------------------------- | ------------------------------ |
| **Cursor**             | ❌          | Elegant interface, good completion experience   | No Agent mode                  |
| **Cline (VS Code)**    | ✅          | Full Agent functionality, can auto-modify files | Requires VS Code               |
| **RooCode (VS Code)**  | ✅          | Agent mode, supports multi-file editing         | Newer, features being improved |
| **Continue (VS Code)** | ✅          | Open source, highly customizable                | More complex configuration     |

## Core Features

### Smart Code Completion

* **Tab Completion**: Press Tab to accept AI suggestions
* **Multi-line Completion**: Supports function-level code generation
* **Context Aware**: Provides suggestions based on project structure

### AI Conversation

* **Ctrl/Cmd + K**: Open command palette
* **Ctrl/Cmd + L**: Sidebar conversation
* **Code Explanation**: Select code and ask AI

### Code Editing

* **Generate Code**: Describe requirements, AI auto-generates
* **Refactoring Suggestions**: Get optimization recommendations
* **Error Fixing**: AI assists in locating and fixing errors

## Keyboard Shortcuts

| Shortcut       | Function               |
| -------------- | ---------------------- |
| `Ctrl/Cmd + K` | AI command palette     |
| `Ctrl/Cmd + L` | AI conversation        |
| `Tab`          | Accept code suggestion |
| `Esc`          | Cancel suggestion      |

## Usage Tips

### 1. Provide Clear Context

```javascript theme={null}
// @context: React component for user authentication
// @requirements: Need to support OAuth2 login
// @constraints: Compatible with NextJS 13+

// AI will generate more accurate code based on this information
```

### 2. Optimize Prompts

```
// Bad prompt
"Fix this function"

// Good prompt
"Fix floating point precision issue in calculateTotal function, ensure amount calculation accurate to 2 decimal places"
```

### 3. Fully Utilize Chat Mode

Although there's no Agent mode, you can:

* Have AI generate complete file content
* Request AI to provide detailed modification instructions
* Use AI for code review and refactoring suggestions

## Troubleshooting

### Connection Timeout

1. Check network connection
2. Confirm API address: `https://api.apiyi.com/v1`
3. Verify API key validity

### Model Not Responding

1. Check account balance
2. Try switching models
3. Restart Cursor

### Poor Code Suggestions

1. Provide more project context
2. Use more specific prompts
3. Try different models

## Best Practices

### Project-Level Configuration

Create `.cursor-settings.json` in project root:

```json theme={null}
{
  "model": "gpt-4.1",
  "temperature": 0.7,
  "contextFiles": ["README.md", "package.json"],
  "rules": [
    "Use TypeScript strict mode",
    "Follow ESLint standards",
    "Add appropriate comments"
  ]
}
```

### Code Standards

Clearly specify code standards in prompts:

* Use TypeScript
* Follow Airbnb standards
* Add JSDoc comments
* Use functional style

### Security Awareness

* Don't include sensitive information in code
* Review AI-generated code
* Verify third-party dependency security

## Integration Workflow

### Git Integration

```bash theme={null}
# AI generates commit message
git add .
# Use Cursor AI to generate descriptive commit message
```

### Test-Driven Development

1. Write test cases first
2. Have AI generate implementation code
3. Run tests to verify
4. Iterate and optimize

### Code Review

Use AI for code review:

```
Please review this code, focusing on:
1. Performance issues
2. Security vulnerabilities
3. Code standards
4. Best practices
```

## Need Agent Mode?

If you need AI to automatically modify multiple files and perform complex refactoring tasks, recommend checking out:

<CardGroup cols={2}>
  <Card title="Cline" icon="bot" href="/scenarios/programming/cline">
    Full-featured AI Agent in VS Code, supports auto-modifying files
  </Card>

  <Card title="RooCode" icon="code" href="https://marketplace.visualstudio.com/items?itemName=roocode.roocode">
    Emerging VS Code AI Agent plugin, supports multi-file editing
  </Card>
</CardGroup>

<Info>
  **Tip**: If you're a Vibe Coding enthusiast and need AI to autonomously complete complex programming tasks, recommend using tools that support Agent mode, or consider purchasing Cursor official membership for full experience.
</Info>

Need more help? Please visit [APIYI Official Website](https://api.apiyi.com) for support.