### Install Dify AI Provider

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Installs the dify-ai-provider package using npm, pnpm, or yarn. This is the first step to integrating Dify AI with the Vercel AI SDK.

```bash
npm install dify-ai-provider

# pnpm
pnpm add dify-ai-provider

# yarn
yarn add dify-ai-provider
```

--------------------------------

### Basic Dify AI Integration with Vercel AI SDK

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Demonstrates how to set up and use the Dify AI provider with the Vercel AI SDK to generate text. It requires setting the DIFY_API_KEY environment variable and providing the Dify application ID. The example shows how to initiate a text generation request and extract conversation and message IDs from the provider metadata.

```typescript
import { generateText } from "ai";
import { difyProvider } from "dify-ai-provider";

process.env.DIFY_API_KEY = "dify-api-key"; // app-...

// Create a Dify provider instance
const dify = difyProvider("dify-application-id", {
  responseMode: "blocking",
});

// Generate text using Dify AI
const { text, providerMetadata } = await generateText({
  model: dify,
  messages: [{ role: "user", content: "Hello, how are you today?" }],
  headers: { "user-id": "test-user" },
});

const { conversationId, messageId } = providerMetadata.difyWorkflowData;
console.log(text);
console.log("conversationId", conversationId);
console.log("messageId", messageId);
```

--------------------------------

### Stream Dify AI Responses

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Demonstrates how to stream responses from Dify AI using the Vercel AI SDK. This allows for a more interactive user experience by processing text chunks as they arrive.

```typescript
import { streamText } from "ai";
import { difyProvider } from "dify-ai-provider";

const dify = difyProvider("dify-application-id");

const stream = streamText({
  model: dify,
  messages: [{ role: "user", content: "Write a short poem about AI." }],
});

// Process text deltas as they arrive
for await (const chunk of stream.textStream) {
  process.stdout.write(chunk);
}
```

--------------------------------

### Configure Self-Hosted Dify Provider

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Explains how to configure the Dify provider for self-hosted Dify instances. This involves specifying the custom baseURL and optionally providing the API key directly in the provider configuration.

```typescript
import { createDifyProvider } from "dify-ai-provider";
const difyProvider = createDifyProvider({
  baseURL: "your-base-url",
});
const dify = difyProvider("dify-application-id", {
  responseMode: "blocking",
  apiKey: "dify-api-key",
});
```

--------------------------------

### Continue Dify Conversation

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Shows how to continue an existing conversation with Dify AI by passing the conversation ID in the headers. This ensures that the AI maintains context from previous interactions.

```typescript
const { text: followUpText } = await generateText({
  model: dify,
  messages: [
    { role: "user", content: "That's great! What can you help me with?" },
  ],
  headers: { "user-id": "test-user", "chat-id": conversationId },
});

console.log("followUpText", followUpText);
```

--------------------------------

### Retrieve Conversation ID from Dify Response

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Illustrates how to extract the Dify conversation ID and message ID from the `providerMetadata` in the `onFinish` callback of Vercel AI SDK calls. This is crucial for maintaining conversation continuity.

```typescript
onFinish: async ({ response, providerMetadata }) => {
  const conversationId = providerMetadata?.difyWorkflowData?.conversationId as string;
  const messageId = providerMetadata?.difyWorkflowData?.messageId as string;
  // Save conversationId for future use
}
```

--------------------------------

### Pass Conversation ID for Continuity in Next.js Chatbot

Source: https://github.com/warmwind/dify-ai-provider/blob/main/README.md

Shows how to pass the user ID and Dify conversation ID in the headers when making `streamText` calls within a Next.js AI Chatbot application to ensure conversation continuity.

```typescript
const stream = createDataStream({
  execute: (dataStream) => {
    const headers = {
      'user-id': session.user.id,
      'chat-id': conversation_id_returned_from_dify
    };

    const result = streamText({
      model: myProvider.languageModel(selectedChatModel),
      headers,
      // ... other options
    });

    // ... rest of implementation
  }
  // ... other options
});
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.