README.md

💎 gem

A more fun way to interact with Gemini.

await ask("am i deserve better?");

Gem is a simple package that allows you to interact with the Gemini API in a more human-friendly way.

It's like talking to a friend, but in a more professional way. It's simplify the process of sending requests to the Gemini API, so you can focus more on the conversation itself. It's even more simpler than other API wrappers.

🚧 This package is still under heavy development. Beware of breaking changes.

🚀 Getting Started

It's only require 3 simple steps to get started.

Install the package

npm install @ezralazuardy/gem

Set the environment variables

GEMINI_API_KEY="<YOUR_GEMINI_API_KEY>"

Ask the Gemini.

import { ask } from "@ezralazuardy/gem";

const response = await ask("what is the weather today?");

console.log(response);

Yup, just like that! Simple, right?

For advanced usage, please read the documentation below.

⚙️ Using Specific Model

By default, the model used is gemini-pro. You can change it by setting the GEMINI_MODEL_ID environment variable.

GEMINI_MODEL_ID="gemini-pro"

Please beware of the pricing and capabilities of each model.

As of v0.0.1, the following models are available:

• gemini-pro (default)
• gemini-1.5-pro-latest

Plase refer to these references: Gemini API Models, Gemini API Pricing.

🚥 Roadmap

The following features are planned to be implemented in the future.

Supported	Method	Version	Usage
🟢	ask	v0.0.1+	Generates text for a given prompt.
🔴	askByStream	-	Streams a text generations.
🔴	object	-	Generates a typed, structured object for a given prompt.
🔴	objectByStream	-	Streams a typed, structured object for a given prompt.

📖 Documentation

A proper documentation is still in progress 🥲

Written in Typescript. Heavily inspired by Vercel AI and Gemini.