Native Image & Video Generation: Nano Banana & Omni Flash — Mastering the Gemini API

Beyond Text: Native Media Generation

Gemini doesn't just understand images and video - it generates and edits them natively, conversationally, in the same API you already use for text. Google groups its native image models under the Nano Banana family.

The Nano Banana Family

from google import genai
import base64

client = genai.Client(vertexai=True, project="my-gcp-project", location="global")

interaction = client.interactions.create(
    model="gemini-3.1-flash-image",
    input="Create a picture of a nano banana dish in a fancy restaurant with a Gemini theme",
)

with open("generated_image.png", "wb") as f:
    f.write(base64.b64decode(interaction.output_image.data))

Veo 3 & Gemini Omni Flash: Native Video Generation

Veo 3 is Google DeepMind's state-of-the-art HD video model on Vertex AI, introducing native audio generation (sound effects, ambient environments, and lip-synced character dialogue) with realistic physics. Alongside it, Gemini Omni Flash (gemini-omni-flash-preview) generates 3–10 second, 720p videos with audio from a text prompt or a still image, and supports conversational, multi-turn editing of the result - describe a change and it re-renders while preserving what you didn't mention.

res1 = client.interactions.create(
    model="gemini-omni-flash-preview",
    input="A woman playing violin outdoors."
)

# Conversational edit - references the previous generation
res2 = client.interactions.create(
    model="gemini-omni-flash-preview",
    previous_interaction_id=res1.id,
    input="Make the violin invisible."
)

All Nano Banana and Omni Flash outputs include an invisible SynthID watermark for programmatic provenance verification.