@ragpipe/plugin-gemini

Google Gemini embedding and generation plugin for ragpipe.

Install

pnpm add ragpipe @ragpipe/plugin-gemini

Usage

import { defineConfig } from "ragpipe";
import { geminiEmbedding, geminiGeneration } from "@ragpipe/plugin-gemini";

export default defineConfig({
  embedding: geminiEmbedding({
    apiKey: process.env.GEMINI_API_KEY ?? "",
    model: "gemini-embedding-001",
  }),
  generation: geminiGeneration({
    apiKey: process.env.GEMINI_API_KEY ?? "",
    model: "gemini-3.1-flash-lite-preview",
    systemPrompt: "Answer based on the provided context.",
  }),
  // ... vectorStore
});

API

geminiEmbedding(options)

Returns an EmbeddingPlugin that calls the Gemini Embedding API.

| Option | Type | Default | Description | |---|---|---|---| | apiKey | string | — | Google AI API key (required) | | model | string | - | Embedding model name (required) |

• Dimensions: 3072
• Rate limit: 800ms between calls (built-in)
• Batch support: embedMany() uses batchEmbedContents for efficient bulk embedding

geminiGeneration(options)

Returns a GenerationPlugin that calls the Gemini Content Generation API.

| Option | Type | Default | Description | |---|---|---|---| | apiKey | string | — | Google AI API key (required) | | model | string | - | Generation model name (required) | | systemPrompt | string | "Answer based on the provided context." | Default system instruction |

• Streaming: generateStream() returns an AsyncIterable<string> via SSE
• History: Pass { history } to include conversation context
• Per-call override: Pass { systemPrompt } at call time to override the default

Get an API Key

1. Go to Google AI Studio
2. Create an API key
3. Set it as GEMINI_API_KEY in your environment

License

MIT