Available Models

Gemini Models (generateContent) -- Recommended

Model	Model ID	Best For
Gemini 3.1 Flash Image	`gemini-3.1-flash-image-preview`	Fast, up to 4K, search grounding
Gemini 3 Pro Image	`gemini-3-pro-image-preview`	Highest quality, advanced reasoning
Gemini 2.5 Flash Image	`gemini-2.5-flash-image`	Fast creative workflows

Imagen Models (predict) -- Dedicated image gen

Model	Model ID	Best For
Imagen 4 Ultra	`imagen-4.0-ultra-generate-001`	Highest quality
Imagen 4 Standard	`imagen-4.0-generate-001`	Standard quality
Imagen 4 Fast	`imagen-4.0-fast-generate-001`	Speed optimized

Key Differences

Feature	Gemini	Imagen
Input	Text + reference images	Text only
Output	1 image per request	Up to 4 per request
Max resolution	4K	2K
Language	Multilingual (French OK)	English only
Image editing	Yes (pass reference images)	No
Aspect ratios	14 options	5 options

Parameters

Gemini Parameters

Parameter	Values
`aspect_ratio`	`"1:1"`, `"9:16"`, `"16:9"`, `"4:5"`, `"5:4"`, `"4:3"`, `"3:4"`, `"3:2"`, `"2:3"`, `"1:4"`, `"4:1"`, `"1:8"`, `"8:1"`, `"21:9"`
`image_size`	`"512px"` (3.1 Flash only), `"1K"`, `"2K"`, `"4K"`

Imagen Parameters

Parameter	Values	Default
`number_of_images`	1-4	4
`image_size`	`"1K"`, `"2K"`	`"1K"`
`aspect_ratio`	`"1:1"`, `"3:4"`, `"4:3"`, `"9:16"`, `"16:9"`	`"1:1"`
`person_generation`	`"dont_allow"`, `"allow_adult"`, `"allow_all"`	`"allow_adult"`

Common Ad Creative Sizes

Format	Aspect Ratio	Use Case
Feed (Meta/Insta)	`1:1`	Posts, carrousel
Story/Reel	`9:16`	Stories, Reels, TikTok
Landscape	`16:9`	YouTube, banners
Portrait (Meta)	`4:5`	Feed optimized

Usage Patterns

Pattern 1: Single Image (Gemini -- recommended)

import os, base64
from google import genai
from google.genai import types

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

response = client.models.generate_content(
    model="gemini-2.0-flash-exp",
    contents="YOUR PROMPT HERE",
    config=types.GenerateContentConfig(
        response_modalities=["IMAGE"],
        image_config=types.ImageConfig(
            aspect_ratio="1:1",
            image_size="2K",
        ),
    ),
)

for i, part in enumerate(response.candidates[0].content.parts):
    if part.inline_data is not None:
        img_data = base64.b64decode(part.inline_data.data)
        with open(f"output_{i}.png", "wb") as f:
            f.write(img_data)
        print(f"Saved output_{i}.png")

Pattern 2: Batch Generation (Imagen -- up to 4 images)

import os, base64
from google import genai
from google.genai import types

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

response = client.models.generate_images(
    model="imagen-4.0-generate-001",
    prompt="YOUR PROMPT IN ENGLISH",
    config=types.GenerateImagesConfig(
        number_of_images=4,
        aspect_ratio="1:1",
        image_size="2K",
        person_generation="allow_adult",
    ),
)

for i, img in enumerate(response.generated_images):
    img.image.save(f"output_{i}.png")
    print(f"Saved output_{i}.png")

Pattern 3: Image Editing (Gemini with reference)

import os
from google import genai
from google.genai import types
from PIL import Image

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

ref_image = types.Part.from_image(Image.open("input.jpg"))

response = client.models.generate_content(
    model="gemini-2.0-flash-exp",
    contents=["Add bold white text overlay saying 'ATELIER GRATUIT'", ref_image],
    config=types.GenerateContentConfig(
        response_modalities=["IMAGE"],
        image_config=types.ImageConfig(aspect_ratio="1:1"),
    ),
)

for part in response.candidates[0].content.parts:
    if part.inline_data is not None:
        import base64
        img_data = base64.b64decode(part.inline_data.data)
        with open("edited.png", "wb") as f:
            f.write(img_data)

Pattern 4: Batch Ad Creatives Generator

Use this pattern to generate multiple ad creatives from a list of prompts and save them organized in folders.

import os, base64, time
from google import genai
from google.genai import types
from pathlib import Path

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

# Configuration
MODEL = "gemini-2.0-flash-exp"
OUTPUT_DIR = Path("creatives")
ASPECT_RATIO = "1:1"  # Change per format: "9:16" for stories, "4:5" for feed
IMAGE_SIZE = "2K"

def generate_creative(prompt: str, filename: str, aspect_ratio: str = ASPECT_RATIO):
    """Generate a single ad creative and save it."""
    response = client.models.generate_content(
        model=MODEL,
        contents=prompt,
        config=types.GenerateContentConfig(
            response_modalities=["IMAGE"],
            image_config=types.ImageConfig(
                aspect_ratio=aspect_ratio,
                image_size=IMAGE_SIZE,
            ),
        ),
    )

    output_path = OUTPUT_DIR / filename
    output_path.parent.mkdir(parents=True, exist_ok=True)

    for part in response.candidates[0].content.parts:
        if part.inline_data is not None:
            img_data = base64.b64decode(part.inline_data.data)
            with open(output_path, "wb") as f:
                f.write(img_data)
            print(f"Saved {output_path}")
            return str(output_path)
    return None

# Example: Generate creatives for multiple ad angles
creatives = [
    {
        "prompt": "YOUR PROMPT 1",
        "filename": "angle1_feed.png",
        "aspect_ratio": "1:1",
    },
    {
        "prompt": "YOUR PROMPT 2",
        "filename": "angle2_story.png",
        "aspect_ratio": "9:16",
    },
]

for creative in creatives:
    generate_creative(
        prompt=creative["prompt"],
        filename=creative["filename"],
        aspect_ratio=creative.get("aspect_ratio", ASPECT_RATIO),
    )
    time.sleep(1)  # Rate limiting

Prompting Tips for Ad Creatives

Structure d'un bon prompt

[STYLE] + [SUJET PRINCIPAL] + [TEXTE A AFFICHER] + [AMBIANCE/COULEURS] + [COMPOSITION]

Exemples de prompts pour Meta Ads

Statement Ad (texte bold):

Minimalist social media ad, dark gradient background (#1a1a2e to #16213e),
bold white sans-serif text centered saying "L'IA EST LA PLUS GRANDE OPPORTUNITE DE 2026",
subtle blue glow effect, clean modern design, no clutter,
professional marketing aesthetic

Person + Context:

Professional photo of a confident young man in his 30s working on a laptop
in a modern bright office, warm lighting, slight smile,
wearing smart casual clothing, shallow depth of field,
cinematic color grading, editorial photography style

Before/After Split:

Split image ad creative: left side shows a tired office worker in grey cubicle
(desaturated, cold tones), right side shows same person working freely
from a beautiful terrace with laptop (warm, vibrant colors),
bold arrow or dividing line in the center,
clean sans-serif text at bottom

Data/Stats Visual:

Clean infographic style ad, dark premium background,
large glowing number "50 000 EUR/mois" in bold gold text,
subtle AI circuit pattern in background,
small text below saying "Deviens Integrateur IA",
modern minimalist design, Meta ads format

REST API (curl) -- Alternative without Python

Gemini

curl -s -X POST \
  "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash-exp:generateContent" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{"parts": [{"text": "YOUR PROMPT"}]}],
    "generationConfig": {
      "responseModalities": ["IMAGE"],
      "imageConfig": {"aspectRatio": "1:1", "imageSize": "2K"}
    }
  }' | python3 -c "
import sys, json, base64
r = json.load(sys.stdin)
for i, part in enumerate(r['candidates'][0]['content']['parts']):
    if 'inlineData' in part:
        with open(f'output_{i}.png', 'wb') as f:
            f.write(base64.b64decode(part['inlineData']['data']))
        print(f'Saved output_{i}.png')
"

Imagen

curl -s -X POST \
  "https://generativelanguage.googleapis.com/v1beta/models/imagen-4.0-generate-001:predict" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "instances": [{"prompt": "YOUR PROMPT IN ENGLISH"}],
    "parameters": {
      "sampleCount": 4,
      "aspectRatio": "1:1",
      "personGeneration": "allow_adult"
    }
  }' | python3 -c "
import sys, json, base64
r = json.load(sys.stdin)
for i, pred in enumerate(r.get('predictions', [])):
    with open(f'imagen_{i}.png', 'wb') as f:
        f.write(base64.b64decode(pred['bytesBase64Encoded']))
    print(f'Saved imagen_{i}.png')
"

Troubleshooting

Error	Solution
`API key not valid`	Check your key at https://aistudio.google.com/apikey
`SAFETY` block	Rephrase prompt, avoid violent/sexual content
`quota exceeded`	Wait or upgrade plan, Gemini free tier = 15 RPM
`model not found`	Check model ID spelling, some are in preview
Imagen returns empty	Prompt may have been filtered; try simpler prompt

google-ai-images

Google AI Image Generation

Setup

1. Install dependencies

2. Get API Key