sap-aicore-provider

SAP AI Core Foundation Models provider plugin for Vercel AI SDK

Overview

This package enables seamless integration of SAP AI Core Foundation Models with the Vercel AI SDK. Use it to access SAP-hosted foundation language models in your Node.js AI applications.

Installation

npm install @ai-foundry/sap-aicore-provider ai

Usage

Using sap-aicore-provider

import { sapAiCore } from '@ai-foundry/sap-aicore-provider';
import { generateText } from 'ai';

const { text } = await generateText({
  model: sapAiCore('sap-aicore/gpt-4o'),
  prompt: 'Hello, how are you?'
});

Configuration

You can configure the SAP AI Core provider using the createSapAiCore function. Below are the most common options:

Basic Example

import { createSapAiCore } from '@ai-foundry/sap-aicore-provider';

const sapAiCore = createSapAiCore({
  deploymentUrl: 'https://your-sap-aicore-instance', // Required: Your SAP AI Core deployment URL
  headers: {
    'Custom-Header': 'value' // Optional: Add any custom headers
  }
});

Using a Token Provider

To automatically fetch an access token before each request, use the tokenProvider option:

const sapAiCore = createSapAiCore({
  deploymentUrl: 'https://your-sap-aicore-instance',
  tokenProvider: {
    accessTokenBaseUrl: 'https://auth.example.com/token', // Auth endpoint
    clientId: 'your-client-id',
    clientSecret: 'your-client-secret',
    cacheMaxAgeMs: 3600000 // Optional: Token cache duration (default 1h)
  }
});

Environment Variables

You can also provide configuration values using environment variables:

• AICORE_DEPLOYMENT_URL
• AICORE_AUTH_URL
• AICORE_CLIENT_ID
• AICORE_CLIENT_SECRET

Tokens are cached for one hour by default to reduce token provider requests.

License

MIT

Contributing

Contributions are welcome! Please open issues or pull requests on GitHub.