@teamflojo/floimg-google

Google AI integration for floimg, providing Gemini image generation, editing, and AI capabilities.

Standing on the Shoulders of Giants

This plugin integrates with Google's Gemini multimodal AI. We provide a consistent FloImg interface while exposing the full power of Google's models.

• Full Gemini power: Native image generation and editing with all Gemini parameters
• Native format: Use Gemini parameters, not a FloImg abstraction
• Their docs are your docs: See Google AI documentation

FloImg orchestrates the workflow (generate -> transform -> save). Gemini does what it does best.

Installation

npm install @teamflojo/floimg-google
# or
pnpm add @teamflojo/floimg-google

Configuration

Set your Google AI API key:

export GOOGLE_AI_API_KEY=your_key_here

Or pass it directly:

import { geminiGenerate } from "@teamflojo/floimg-google";

const generator = geminiGenerate({ apiKey: "your_key_here" });

Features

Image Generation (Gemini)

Generate images using Gemini's native image generation:

import { createClient } from "@teamflojo/floimg";
import { geminiGenerate } from "@teamflojo/floimg-google";

const floimg = createClient();
floimg.registerGenerator(geminiGenerate());

const image = await floimg.generate({
  generator: "gemini-generate",
  params: {
    prompt: "A serene mountain landscape at sunset",
    aspectRatio: "16:9",
    imageSize: "2K",
  },
});

Parameters:

| Parameter | Type | Default | Description | | ------------------- | -------------------- | ------------------------ | ------------------------------------------------- | | prompt | string | (required) | Image description | | prePrompt | string | "Generate an image..." | Instructions prepended to prompt | | model | string | "gemini-2.5-flash-image" | Gemini model to use | | aspectRatio | string | "1:1" | Output aspect ratio (1:1, 16:9, 9:16, etc.) | | imageSize | "1K" | "2K" | "4K" | "1K" | Output resolution | | groundingWithSearch | boolean | false | Enable Google Search grounding for real-time data | | enhancePrompt | boolean | false | Auto-enhance prompt using best practices | | referenceImages | ImageBlob[] | undefined | Up to 14 reference images for style/consistency |

Image Editing (Gemini Transform)

Edit existing images using Gemini's multimodal capabilities:

import { geminiTransform } from "@teamflojo/floimg-google";

floimg.registerTransformProvider(geminiTransform());

const edited = await floimg.transform({
  blob: inputImage,
  op: "edit",
  provider: "gemini-transform",
  params: {
    prompt: "Make the sky more vibrant and add clouds",
    enhancePrompt: true,
  },
});

Transform Parameters:

| Parameter | Type | Default | Description | | --------------- | ------- | --------- | ---------------------------------------- | | prompt | string | (req.) | Describe the edits to make | | prePrompt | string | "Edit..." | Instructions prepended to prompt | | enhancePrompt | boolean | false | Auto-enhance prompt using best practices | | referenceImages | array | [] | Up to 13 additional reference images |

Prompt Enhancement

The enhancePrompt option automatically expands your prompts based on Google's image generation best practices. It detects the type of image you're creating and applies appropriate enhancements.

// Without enhancement
const image = await floimg.generate({
  generator: "gemini-generate",
  params: {
    prompt: "photo of a mountain",
  },
});
// Prompt sent: "photo of a mountain"

// With enhancement
const image = await floimg.generate({
  generator: "gemini-generate",
  params: {
    prompt: "photo of a mountain",
    enhancePrompt: true,
  },
});
// Prompt sent: "A photorealistic photo of a mountain. The scene is
// illuminated by soft natural light. Captured with professional
// photography equipment, sharp focus, high detail"

Detected Prompt Types:

| Type | Keywords | Enhancement Style | | -------------- | --------------------------------- | ----------------------------------- | | photorealistic | photo, realistic, camera, lens | Adds lighting, camera details | | portrait | portrait, headshot, face, person | Photography terms, focus details | | landscape | landscape, mountain, ocean, sky | Lighting, atmosphere | | illustration | illustration, cartoon, anime, art | Style, color palette | | logo | logo, brand, icon, emblem | Modern, professional, scalable | | product | product, e-commerce, mockup | Studio lighting, neutral background | | minimalist | minimalist, simple, clean | Negative space, subtle lighting | | edit | change, modify, remove, add | Preservation instructions |

Using Enhancement Utilities Directly:

import { enhancePrompt, detectPromptType, isPromptDetailed } from "@teamflojo/floimg-google";

// Detect prompt type
const type = detectPromptType("photo of a sunset"); // "photorealistic"

// Check if prompt needs enhancement
const needsEnhancement = !isPromptDetailed("cat"); // true

// Enhance a prompt manually
const enhanced = enhancePrompt("photo of a mountain", "generate");

Vision Analysis (Gemini Vision)

Analyze images using Gemini's multimodal understanding:

import { geminiVision } from "@teamflojo/floimg-google";

floimg.registerVisionProvider(geminiVision());

const analysis = await floimg.analyze({
  blob: inputImage,
  provider: "gemini-vision",
  params: {
    prompt: "Describe this image in detail",
  },
});

Text Generation (Gemini Text)

Generate text for image workflows:

import { geminiText } from "@teamflojo/floimg-google";

floimg.registerTextProvider(geminiText());

const result = await floimg.text({
  provider: "gemini-text",
  params: {
    prompt: "Write a caption for this sunset photo",
    context: analysis.content,
  },
});

Available Models

| Model | Description | | -------------------------- | ------------------------------------------- | | gemini-2.5-flash-image | Fast, high-volume, low-latency (default) | | gemini-3-pro-image-preview | Professional quality, better text rendering |

License

MIT