Cloudflare Workers AI | AI SDK 한국어

workers-ai-provider is a community provider that allows you to use Cloudflare's Workers AI models with the AI SDK.

Setup

The Cloudflare Workers AI provider is available in the workers-ai-provider module. You can install it with:

pnpm

npm

yarn

bun

pnpm add workers-ai-provider

npm install workers-ai-provider

yarn add workers-ai-provider

bun add workers-ai-provider

pnpmnpmyarnbun

pnpm add workers-ai-provider

Then, setup an AI binding in your Cloudflare Workers project wrangler.toml file:

[ai]
binding = "AI"

Provider Instance

To create a workersai provider instance, use the createWorkersAI function, passing in the AI binding as an option:

import { createWorkersAI } from 'workers-ai-provider';

const workersai = createWorkersAI({ binding: env.AI });

Language Models

To create a model instance, call the provider instance and specify the model you would like to use as the first argument. You can also pass additional settings in the second argument:

import { createWorkersAI } from 'workers-ai-provider';

const workersai = createWorkersAI({ binding: env.AI });
const model = workersai('@cf/meta/llama-3.1-8b-instruct', {
  // additional settings
  safePrompt: true,
});

You can use the following optional settings to customize:

safePrompt boolean

Whether to inject a safety prompt before all conversations. Defaults to false

Examples

You can use Cloudflare Workers AI language models to generate text with the generateText or streamText function:

`generateText`

import { createWorkersAI } from 'workers-ai-provider';
import { generateText } from 'ai';

type Env = {
  AI: Ai;
};

export default {
  async fetch(_: Request, env: Env) {
    const workersai = createWorkersAI({ binding: env.AI });
    const result = await generateText({
      model: workersai('@cf/meta/llama-2-7b-chat-int8'),
      prompt: 'Write a 50-word essay about hello world.',
    });

    return new Response(result.text);
  },
};

`streamText`

import { createWorkersAI } from 'workers-ai-provider';
import { createTextStreamResponse, streamText, toTextStream } from 'ai';

type Env = {
  AI: Ai;
};

export default {
  async fetch(_: Request, env: Env) {
    const workersai = createWorkersAI({ binding: env.AI });
    const result = streamText({
      model: workersai('@cf/meta/llama-2-7b-chat-int8'),
      prompt: 'Write a 50-word essay about hello world.',
    });

    return createTextStreamResponse({
      stream: toTextStream({ stream: result.stream }),
      headers: {
        // add these headers to ensure that the
        // response is chunked and streamed
        'Content-Type': 'text/x-unknown',
        'content-encoding': 'identity',
        'transfer-encoding': 'chunked',
      },
    });
  },
};

Structured Output

Some Cloudflare Workers AI language models also support structured data generation with Output:

import { createWorkersAI } from 'workers-ai-provider';
import { generateText, Output } from 'ai';
import { z } from 'zod';

type Env = {
  AI: Ai;
};

export default {
  async fetch(_: Request, env: Env) {
    const workersai = createWorkersAI({ binding: env.AI });
    const result = await generateText({
      model: workersai('@cf/meta/llama-3.1-8b-instruct'),
      prompt: 'Generate a Lasagna recipe',
      output: Output.object({
        schema: z.object({
          recipe: z.object({
            ingredients: z.array(z.string()),
            description: z.string(),
          }),
        }),
      }),
    });

    return Response.json(result.output);
  },
};