@flowrag/provider-gemini

Gemini AI provider for FlowRAG - embeddings and entity extraction.

Installation

``bash
npm install @flowrag/provider-gemini
`

Usage

$3

`typescript
import { GeminiEmbedder } from '@flowrag/provider-gemini';

const embedder = new GeminiEmbedder({
apiKey: 'your-gemini-api-key', // or set GEMINI_API_KEY env var
model: 'text-embedding-004', // optional, default
});

// Single embedding
const embedding = await embedder.embed('Hello world');

// Batch embeddings
const embeddings = await embedder.embedBatch(['Hello', 'World']);
`

$3

`typescript
import { GeminiExtractor } from '@flowrag/provider-gemini';
import { defineSchema } from '@flowrag/core';

const extractor = new GeminiExtractor({
apiKey: 'your-gemini-api-key', // or set GEMINI_API_KEY env var
model: 'gemini-2.0-flash-exp', // optional, default
temperature: 0.1, // optional, default
});

const schema = defineSchema({
entityTypes: ['SERVICE', 'DATABASE'],
relationTypes: ['USES', 'PRODUCES'],
});

const result = await extractor.extractEntities(
'ServiceA connects to DatabaseB',
['ServiceC'], // known entities
schema
);
`

Environment Variables

`bash
GEMINI_API_KEY=your-api-key
``

License

MIT