vision-agents-plugins-aws 0.2.0

AWS (Bedrock) LLM integration for Vision Agents

AWS Plugin for Vision Agents

AWS (Bedrock) LLM integration for Vision Agents framework with support for both standard and realtime interactions.

Installation

pip install vision-agents-plugins-aws

Usage

Standard LLM Usage

This example shows how to use qwen3 on bedrock for the LLM.

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="Friendly AI"),
    instructions="Be nice to the user",
    llm=aws.LLM(model="qwen.qwen3-32b-v1:0"),
    tts=cartesia.TTS(),
    stt=deepgram.STT(),
    turn_detection=smart_turn.TurnDetection(buffer_duration=2.0, confidence_threshold=0.5),
)

The full example is available in example/aws_qwen_example.py

Realtime Audio Usage

Nova Sonic audio realtime STS is also supported:

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="Story Teller AI"),
    instructions="Tell a story suitable for a 7 year old about a dragon and a princess",
    llm=aws.Realtime(),
)

Function Calling

Standard LLM (aws.LLM)

The standard LLM implementation fully supports function calling. Register functions using the @llm.register_function decorator:

from vision_agents.plugins import aws

llm = aws.LLM(
    model="qwen.qwen3-32b-v1:0",
    region_name="us-east-1"
)

@llm.register_function(
    name="get_weather",
    description="Get the current weather for a given city"
)
def get_weather(city: str) -> dict:
    """Get weather information for a city."""
    return {
        "city": city,
        "temperature": 72,
        "condition": "Sunny"
    }

Realtime (aws.Realtime)

The Realtime implementation fully supports function calling with AWS Nova Sonic. Register functions using the @llm.register_function decorator:

from vision_agents.plugins import aws

llm = aws.Realtime(
    model="amazon.nova-sonic-v1:0",
    region_name="us-east-1"
)

@llm.register_function(
    name="get_weather",
    description="Get the current weather for a given city"
)
def get_weather(city: str) -> dict:
    """Get weather information for a city."""
    return {
        "city": city,
        "temperature": 72,
        "condition": "Sunny"
    }

# The function will be automatically called when the model decides to use it

See example/aws_realtime_function_calling_example.py for a complete example.

Running the examples

Create a .env file, or cp .env.example to .env and fill in

STREAM_API_KEY=your_stream_api_key_here
STREAM_API_SECRET=your_stream_api_secret_here

AWS_BEARER_TOKEN_BEDROCK=
AWS_ACCESS_KEY_ID=
AWS_SECRET_ACCESS_KEY=

FAL_KEY=
CARTESIA_API_KEY=
DEEPGRAM_API_KEY=