Image Tool with Vertex AI

Use an OpenAI agent to call a Gemini image-generation agent through a custom tool.

Use a custom tool to call Gemini’s current image-generation path. The OpenAI agent chooses the prompt; the Gemini agent returns image bytes through ToolResult.

Current example

Install the dependencies in your Python environment:

uv pip install -U agno google-genai openai pillow

Enable the API and billing in your Google Cloud project, then configure Application Default Credentials:

gcloud auth application-default login
export GOOGLE_CLOUD_PROJECT="your-project-id"
export GOOGLE_CLOUD_LOCATION="global"
export OPENAI_API_KEY="your-api-key"

The image model supports the global endpoint; see Gemini 3.1 Flash Image on Vertex AI.

Save this as imagen_tool_advanced.py:

imagen_tool_advanced.py
import os
from io import BytesIO
from pathlib import Path

from agno.agent import Agent
from agno.models.google import Gemini
from agno.models.openai import OpenAIChat
from agno.tools.function import ToolResult
from PIL import Image as PILImage

image_agent = Agent(
    model=Gemini(
        id="gemini-3.1-flash-image",
        response_modalities=["TEXT", "IMAGE"],
        vertexai=True,
        project_id=os.environ["GOOGLE_CLOUD_PROJECT"],
        location=os.environ.get("GOOGLE_CLOUD_LOCATION", "global"),
    ),
)


def generate_image(prompt: str) -> ToolResult:
    """Generate an image from a detailed visual description."""
    result = image_agent.run(prompt)
    if not result.images:
        raise RuntimeError("Gemini returned no image")
    return ToolResult(content=str(result.content or "Image generated"), images=result.images)


agent = Agent(
    model=OpenAIChat(id="gpt-5.2"),
    tools=[generate_image],
    instructions="Use generate_image for image requests.",
)
response = agent.run("Create a painting of a floating city in the clouds at sunset.")
if not response.images:
    raise RuntimeError("No generated image was returned")
output_dir = Path("tmp/generated-images")
output_dir.mkdir(parents=True, exist_ok=True)
for index, image in enumerate(response.images):
    if not image.content:
        raise RuntimeError("Generated image has no bytes")
    output_path = output_dir / f"image_{index}.png"
    with PILImage.open(BytesIO(image.content)) as decoded:
        decoded.save(output_path)
    print(output_path)

Run python imagen_tool_advanced.py. This saves the generated images as PNG files. The nested Gemini call uses generate_content; substituting a Gemini model ID into GeminiTools.generate_image() would still call the Imagen generate_images endpoint.

Historical Imagen example

The original example below is retained for comparison and cannot run as written. It uses a retired Imagen model, the removed Agent.run_response accessor, and passes raw bytes to a base64 decoder. Use the complete current example above. Vertex Imagen 4 was discontinued on June 30, 2026; Gemini API Imagen models have a separate retirement schedule.

"""Example: Using the GeminiTools Toolkit for Image Generation

An Agent using the Gemini image generation tool.

Make sure to set the Vertex AI credentials. Here's the authentication guide: https://cloud.google.com/sdk/docs/initializing

Run `uv pip install google-genai agno` to install the required packages.
"""

from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.models.gemini import GeminiTools
from agno.utils.media import save_base64_data

agent = Agent(
    model=OpenAIChat(id="gpt-4o"),
    tools=[
        GeminiTools(
            image_generation_model="imagen-4.0-generate-preview-05-20", vertexai=True
        )
    ],
)

agent.print_response(
    "Cinematic a visual shot using a stabilized drone flying dynamically alongside a pod of immense baleen whales as they breach spectacularly in deep offshore waters. The camera maintains a close, dramatic perspective as these colossal creatures launch themselves skyward from the dark blue ocean, creating enormous splashes and showering cascades of water droplets that catch the sunlight. In the background, misty, fjord-like coastlines with dense coniferous forests provide context. The focus expertly tracks the whales, capturing their surprising agility, immense power, and inherent grace. The color palette features the deep blues and greens of the ocean, the brilliant white spray, the dark grey skin of the whales, and the muted tones of the distant wild coastline, conveying the thrilling magnificence of marine megafauna."
)

response = agent.run_response
if response and response.images:
    save_base64_data(str(response.images[0].content), "tmp/baleen_whale.png")

"""
Example prompts to try:
- A horizontally oriented rectangular stamp features the Mission District's vibrant culture, portrayed in shades of warm terracotta orange using an etching style. The scene might depict a sun-drenched street like Valencia or Mission Street, lined with a mix of Victorian buildings and newer structures.
- Painterly landscape featuring a simple, isolated wooden cabin nestled amongst tall pine trees on the shore of a calm, reflective lake.
- Filmed cinematically from the driver's seat, offering a clear profile view of the young passenger on the front seat with striking red hair.
- A pile of books seen from above. The topmost book contains a watercolor illustration of a bird. VERTEX AI is written in bold letters on the book.
"""