Tracing OpenAI

MLflow Tracing provides automatic tracing capability for OpenAI. By enabling auto tracing for OpenAI by calling the mlflow.openai.autolog function, MLflow will capture traces for LLM invocation and log them to the active MLflow Experiment.

MLflow trace automatically captures the following information about OpenAI calls:

• Prompts and completion responses
• Latencies
• Model name
• Additional metadata such as temperature, max_tokens, if specified.
• Function calling if returned in the response
• Any exception if raised

Prerequisites

To use MLflow Tracing with OpenAI, you need to install MLflow and the OpenAI SDK.

Development

For development environments, install the full MLflow package with Databricks extras and openai:

Bash

pip install --upgrade "mlflow[databricks]>=3.1" openai

The full mlflow[databricks] package includes all features for local development and experimentation on Databricks.

Production

For production deployments, install mlflow-tracing and openai:

Bash

pip install --upgrade mlflow-tracing openai

The mlflow-tracing package is optimized for production use.

note

MLflow 3 is highly recommended for the best tracing experience with OpenAI.

Before running the examples, you'll need to configure your environment:

For users outside Databricks notebooks: Set your Databricks environment variables:

Bash

export DATABRICKS_HOST="https://your-workspace.cloud.databricks.com"
export DATABRICKS_TOKEN="your-personal-access-token"

For users inside Databricks notebooks: These credentials are automatically set for you.

API Keys: Ensure your OpenAI API key is set:

Bash

export OPENAI_API_KEY="your-openai-api-key"

Supported APIs

MLflow supports automatic tracing for the following OpenAI APIs:

Chat Completion	Embeddings	Function Calling	Structured Outputs	Streaming	Async	Image	Audio
✅	✅	✅	✅ (*1)	✅ (*2)	✅ (*1)

(*1) Streaming support was added in MLflow 2.15.0.

(*2) Async and structured output supported were added in MLflow 2.21.0.

To request support for additional APIs, please open a feature request on GitHub.

Basic Example

Python

import openai
import mlflow
import os

# Ensure your OPENAI_API_KEY is set in your environment
# os.environ["OPENAI_API_KEY"] = "your-openai-api-key" # Uncomment and set if not globally configured

# Enable auto-tracing for OpenAI
mlflow.openai.autolog()

# Set up MLflow tracking to Databricks
mlflow.set_tracking_uri("databricks")
mlflow.set_experiment("/Shared/openai-tracing-demo")

openai_client = openai.OpenAI()

messages = [
    {
        "role": "user",
        "content": "What is the capital of France?",
    }
]

response = openai_client.chat.completions.create(
    model="gpt-4o-mini",
    messages=messages,
    temperature=0.1,
    max_tokens=100,
)

Streaming

MLflow Tracing supports streaming API of the OpenAI SDK. With the same set up of auto tracing, MLflow automatically traces the streaming response and render the concatenated output in the span UI. The actual chunks in the response stream can be found in the Event tab as well.

Python

import openai
import mlflow
import os

# Ensure your OPENAI_API_KEY is set in your environment
# os.environ["OPENAI_API_KEY"] = "your-openai-api-key" # Uncomment and set if not globally configured

# Enable trace logging
mlflow.openai.autolog()

# Set up MLflow tracking to Databricks if not already configured
# mlflow.set_tracking_uri("databricks")
# mlflow.set_experiment("/Shared/openai-streaming-demo")

client = openai.OpenAI()

stream = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {"role": "user", "content": "How fast would a glass of water freeze on Titan?"}
    ],
    stream=True,  # Enable streaming response
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

Async

MLflow Tracing supports asynchronous API of the OpenAI SDK since MLflow 2.21.0. The usage is same as the synchronous API.

Python

import openai
import mlflow
import os

# Ensure your OPENAI_API_KEY is set in your environment
# os.environ["OPENAI_API_KEY"] = "your-openai-api-key" # Uncomment and set if not globally configured

# Enable trace logging
mlflow.openai.autolog()

# Set up MLflow tracking to Databricks if not already configured
# mlflow.set_tracking_uri("databricks")
# mlflow.set_experiment("/Shared/openai-async-demo")

client = openai.AsyncOpenAI()

response = await client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[
        {"role": "user", "content": "How fast would a glass of water freeze on Titan?"}
    ],
    # Async streaming is also supported
    # stream=True
)

Function Calling

MLflow Tracing automatically captures function calling response from OpenAI models. The function instruction in the response will be highlighted in the trace UI. Moreover, you can annotate the tool function with the @mlflow.trace decorator to create a span for the tool execution.

The following example implements a simple function calling agent using OpenAI Function Calling and MLflow Tracing for OpenAI.

Python

import json
from openai import OpenAI
import mlflow
from mlflow.entities import SpanType
import os

# Ensure your OPENAI_API_KEY is set in your environment
# os.environ["OPENAI_API_KEY"] = "your-openai-api-key" # Uncomment and set if not globally configured

# Set up MLflow tracking to Databricks if not already configured
# mlflow.set_tracking_uri("databricks")
# mlflow.set_experiment("/Shared/openai-function-agent-demo")

# Assuming autolog is enabled globally or called earlier
# mlflow.openai.autolog()

client = OpenAI()


# Define the tool function. Decorate it with `@mlflow.trace` to create a span for its execution.
@mlflow.trace(span_type=SpanType.TOOL)
def get_weather(city: str) -> str:
    if city == "Tokyo":
        return "sunny"
    elif city == "Paris":
        return "rainy"
    return "unknown"


tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "parameters": {
                "type": "object",
                "properties": {"city": {"type": "string"}},
            },
        },
    }
]

_tool_functions = {"get_weather": get_weather}


# Define a simple tool calling agent
@mlflow.trace(span_type=SpanType.AGENT)
def run_tool_agent(question: str):
    messages = [{"role": "user", "content": question}]

    # Invoke the model with the given question and available tools
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=messages,
        tools=tools,
    )
    ai_msg = response.choices[0].message
    messages.append(ai_msg)

    # If the model request tool call(s), invoke the function with the specified arguments
    if tool_calls := ai_msg.tool_calls:
        for tool_call in tool_calls:
            function_name = tool_call.function.name
            if tool_func := _tool_functions.get(function_name):
                args = json.loads(tool_call.function.arguments)
                tool_result = tool_func(**args)
            else:
                raise RuntimeError("An invalid tool is returned from the assistant!")

            messages.append(
                {
                    "role": "tool",
                    "tool_call_id": tool_call.id,
                    "content": tool_result,
                }
            )

        # Sent the tool results to the model and get a new response
        response = client.chat.completions.create(
            model="gpt-4o-mini", messages=messages
        )

    return response.choices[0].message.content


# Run the tool calling agent
question = "What's the weather like in Paris today?"
answer = run_tool_agent(question)

Disable auto-tracing

Auto tracing for OpenAI can be disabled globally by calling mlflow.openai.autolog(disable=True) or mlflow.autolog(disable=True).