Gemini block: Add support for tool code execution (tool use)

[Erol444]

What does this PR do?

Add support for Gemini's code execution (tool use)

Type of Change

• New feature (non-breaking change that adds functionality)

Testing

• I have tested this change locally

• I have added/updated tests for this change

• Tested locally both with gemini 3.0 flash and pro, they both support code execution

• Issue: if we do rf_key:roboflow (api key passthrough), seems like our passthroguh server strips away the code execution code, and it won't do code execution

Tested with code below.

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code where necessary, particularly in hard-to-understand areas
• My changes generate no new warnings or errors
• I have updated the documentation accordingly (if applicable)

Additional Context

Test code

import cv2
import traceback
from unittest.mock import MagicMock
from inference.core.workflows.core_steps.models.foundation.google_gemini.v3 import GoogleGeminiBlockV3
from inference.core.workflows.execution_engine.entities.base import WorkflowImageData

# Configuration
google_api_key = "GEMINI_API_KEY"
roboflow_api_key = "ROBOFLOW_API_KEY"
img_path = "sc-highway.png"
prompt = "what's the license plate number of the third (furthest away) car? return just the license plate num."

# Load and prepare image
img = cv2.imread(img_path)
if img is None:
    raise FileNotFoundError(f"Could not load image at {img_path}")
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

# Initialize Block
block = GoogleGeminiBlockV3(
    model_manager=MagicMock(),
    api_key=roboflow_api_key
)

image_data = WorkflowImageData(
    parent_metadata=MagicMock(),
    numpy_image=img
)

try:
    result = block.run(
        images=[image_data],
        task_type="unconstrained",
        prompt=prompt,
        output_structure=None,
        classes=None,
        model_version="gemini-3-flash-preview",
        max_tokens=8192,
        temperature=0.0,
        thinking_level="high",
        max_concurrent_requests=1,
        google_code_execution=True, # <-- Added this
        api_key=google_api_key
    )

    print("\n✅ Execution Successful!")
    print("Response:", result[0]['output'])

except Exception:
    print("\n❌ Execution Failed")
    traceback.print_exc()

And test img:

[PawelPeczek-Roboflow]

I am approving the PR, but there is a problem at this very moment with GH UI - namely approve button is missing 😆 I am exercising admin rights to merge