AgentOps provides seamless integration with IBM Watsonx.ai Python SDK, allowing you to track and analyze all your Watsonx.ai model interactions automatically.

​
Installation

Make sure to call agentops.init before calling any openai, cohere, crew, etc models.

pip install agentops ibm-watsonx-ai

​
Basic Usage

Initialize AgentOps at the beginning of your application to automatically track all IBM Watsonx.ai API calls:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

# Project ID
project_id = "<YOUR_PROJECT_ID>"

# Create a model instance
model = ModelInference(
    model_id="meta-llama/llama-3-3-70b-instruct",
    credentials=credentials,
    project_id=project_id
)

# Make a completion request - AgentOps will track it automatically
response = model.generate_text("What is artificial intelligence?")
print(f"Generated Text:\n{response}")

# Don't forget to close connection when done
model.close_persistent_connection()

​
Chat Completion Example

Using the Watsonx.ai SDK for chat-based interactions:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

# Project ID
project_id = "<YOUR_PROJECT_ID>"

# Create model for chat
chat_model = ModelInference(
    model_id="meta-llama/llama-3-3-70b-instruct",
    credentials=credentials,
    project_id=project_id
)

# Format messages for chat method
messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "What are the three laws of robotics?"}
]

# Get chat response
chat_response = chat_model.chat(messages)
print(f"Chat Response:\n{chat_response['choices'][0]['message']['content']}")

# Close connection
chat_model.close_persistent_connection()

​
Advanced Examples

​
Streaming Example

AgentOps automatically tracks streaming completions:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

project_id = "<YOUR_PROJECT_ID>"

# Create model for streaming
model = ModelInference(
    model_id="google/flan-ul2",
    credentials=credentials,
    project_id=project_id
)

# Text streaming
stream_response = model.generate_text_stream(
    prompt="Write a short poem about artificial intelligence."
)

print("Streaming Response:")
for chunk in stream_response:
    if isinstance(chunk, str):
        print(chunk, end="", flush=True)

# Close connection
model.close_persistent_connection()

​
Chat Streaming Example

Stream responses from chat-based models:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

project_id = "<YOUR_PROJECT_ID>"

# Create model for chat streaming
chat_model = ModelInference(
    model_id="meta-llama/llama-3-3-70b-instruct",
    credentials=credentials,
    project_id=project_id
)

# Format messages for chat method
chat_messages = [
    {"role": "system", "content": "You are a concise assistant."},
    {"role": "user", "content": "Explain the concept of photosynthesis in one sentence."}
]

# Get streaming chat response
chat_stream = chat_model.chat_stream(messages=chat_messages)

print("Chat Stream Response:")
for chunk in chat_stream:
    if chunk and 'choices' in chunk and chunk['choices']:
        delta = chunk['choices'][0].get('delta', {})
        content = delta.get('content')
        if content:
            print(content, end="", flush=True)

# Close connection
chat_model.close_persistent_connection()

​
With Parameters Example

Using the Watsonx.ai SDK with specific generation parameters:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference
from ibm_watsonx_ai.foundation_models.schema import TextGenParameters

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

project_id = "<YOUR_PROJECT_ID>"

# Create model with parameters
model = ModelInference(
    model_id="google/flan-ul2",
    credentials=credentials,
    project_id=project_id,
    params=TextGenParameters(
        decoding_method="sample",
        max_new_tokens=100,
        min_new_tokens=10,
        temperature=0.7,
        top_p=0.9,
        top_k=50,
        repetition_penalty=1.2,
        random_seed=42
    )
)

# Generate text
response = model.generate_text("Explain quantum computing in simple terms.")
print(f"Response:\n{response}")

# Close connection
model.close_persistent_connection()

​
Tokenization Example

Tokenize text with the Watsonx.ai SDK:

import agentops
from ibm_watsonx_ai import Credentials
from ibm_watsonx_ai.foundation_models import ModelInference

# Initialize AgentOps
agentops.init(api_key="<INSERT YOUR API KEY HERE>")

# Initialize credentials
credentials = Credentials(
    url="<YOUR_IBM_URL>",
    api_key="<YOUR_IBM_API_KEY>",
)

project_id = "<YOUR_PROJECT_ID>"

# Create model
model = ModelInference(
    model_id="google/flan-ul2",
    credentials=credentials,
    project_id=project_id
)

# Tokenize text
text_to_tokenize = "Hello, how are you today?"
tokens = model.tokenize(text_to_tokenize)
print(f"Tokenization Result:\n{tokens}")

# Close connection
model.close_persistent_connection()

Read more about environment variables in Advanced Configuration

​
Additional Resources