@qualve/googleai

Google Gemini provider for Qualve LLM tasks.

Setup

Requires Node.js v23+.

npm install @qualve/googleai

Set the GEMINI_API_KEY environment variable (or add it to .env). Get a key at https://aistudio.google.com/api-keys.

Usage

import "@qualve/googleai";

Importing the package registers the gemini provider with the Qualve task system.

Then use llm: "gemini" in your task definitions:

export default {
	type: "llm",
	llm: "gemini",
	system: "You are a helpful assistant.",
	prompt: "Summarize this data.",
	input: [{ name: "data", schema: mySchema }],
	output: { name: "summary", schema: summarySchema },
};

Models

| Model | Context window | Max output | | --- | --- | --- | | gemini-3.1-pro-preview (default) | 1,048,576 | 65,536 | | gemini-3.1-flash-preview | 1,048,576 | 65,536 | | gemini-3.1-flash-lite-preview | 1,048,576 | 65,536 |

Capabilities

| Capability | Supported | | --- | --- | | Structured output (JSON schema) | Yes | | Thinking levels | Yes (minimal, low, medium, high*) | | Web search | Yes (pro models only) | | Token counting | Yes |

* Default