DEV Community

Stream profile image Sarah Lindauer
Sarah Lindauer for Stream

Posted on • Originally published at getstream.io

Build a Vision AI Agent with Gemini 3 in < 3 Minutes

#visionai #gemini #visionagent

Stream released support for Google's new Gemini 3 models inside Vision Agents — the open-source Python framework for building real-time voice and video AI applications.

In this 3-minute video demo, you'll see how to spin up a fully functional vision-enabled voice agent that can see your screen (or webcam), reason with Gemini 3 Pro Preview, and talk back to you naturally, all in pure Python.

What You'll Learn

  • Install Vision Agents + the new Gemini plugin

  • Use gemini-3-pro-preview as your LLM with a single line

  • Build a live video-call agent that can see and describe anything on your screen in real time

  • Customize reasoning depth (low/high thinking level)

Get Started in 60 Seconds

1. Create a fresh project (we recommend uv).

# Initialize a new Python project
uv init

# Activate your environment
uv venv && source .venv/bin/activate
Enter fullscreen mode Exit fullscreen mode

2. Install Vision Agents + project plugins.

# Install Vision Agents
uv add vision-agents

# Install required plugins
uv add "vision-agents[getstream, gemini, elevenlabs, deepgram, smart-turn]"
Enter fullscreen mode Exit fullscreen mode

You'll also need:

Minimal Working Example

Rename your [main.py](http://main.py) to gemini_vision_demo.py and replace its content with this sample code.

import asyncio
import logging

from dotenv import load_dotenv
from vision_agents.core import User, Agent, cli
from vision_agents.core.agents import AgentLauncher
from vision_agents.plugins import elevenlabs, getstream, smart_turn, gemini, deepgram

logger = logging.getLogger(__name__)

load_dotenv()

async def create_agent(**kwargs) -> Agent:
    """Create the agent with Inworld AI TTS."""
    agent = Agent(
        edge=getstream.Edge(),
        agent_user=User(name="Friendly AI", id="agent"),
        instructions="You are a friendly AI assistant powered by Gemini 3. You are able to answer questions and help with tasks. You carefully observe a users' camera feed and respond to their questions and tasks.",
        tts=elevenlabs.TTS(),
        stt=deepgram.STT(),
        # Gemini 3 model
        llm=gemini.LLM("gemini-3-pro-preview"),
        turn_detection=smart_turn.TurnDetection(),
    )
    return agent

async def join_call(agent: Agent, call_type: str, call_id: str, **kwargs) -> None:
    """Join the call and start the agent."""
    # Ensure the agent user is created
    await agent.create_user()
    # Create a call
    call = await agent.create_call(call_type, call_id)

    logger.info("🤖 Starting Inworld AI Agent...")

    # Have the agent join the call/room
    with await agent.join(call):
        logger.info("Joining call")
        logger.info("LLM ready")

        await asyncio.sleep(5)
        await agent.llm.simple_response(text="Describe what you currently see")
        await agent.finish()  # Run till the call ends

if __name__ == "__main__":
    cli(AgentLauncher(create_agent=create_agent, join_call=join_call))
Enter fullscreen mode Exit fullscreen mode

Run it:

uv run gemini_vision_demo.py

A browser tab opens with a Stream Video call. Click "Join call", grant camera/mic/screen permissions, and say something like:

"Okay, I'm going to share my screen — tell me what you see!"

Gemini 3 will instantly analyze your screen and respond with surprisingly detailed descriptions, all in a natural spoken voice.

Links & Resources

Gemini 3 brings better reasoning and multimodal understanding, and Vision Agents makes it simple to turn that power into interactive voice/video experiences. No React, no WebRTC boilerplate, just Python.

Try it today! 🚀

Top comments (0)