Skip to content
OpenObserve
Get Demo Star
OpenObserve
Get Demo Star
Open Menu Close Menu

OpenAI Assistants API → OpenObserve

Automatically capture thread creation, message handling, and run execution for every OpenAI Assistants API workflow in your Python application using the standard OpenAI instrumentor.

Note

The Assistants API is deprecated by OpenAI in favour of the Responses API. It continues to function and is fully instrumented by the OpenAI instrumentor.

Prerequisites

  • Python 3.8+
  • An OpenObserve account (cloud or self-hosted)
  • Your OpenObserve organisation ID and Base64-encoded auth token
  • An OpenAI API key

Installation

pip install openobserve-telemetry-sdk openinference-instrumentation-openai openai python-dotenv

Configuration

Create a .env file in your project root:

OPENOBSERVE_URL=https://api.openobserve.ai/
OPENOBSERVE_ORG=your_org_id
OPENOBSERVE_AUTH_TOKEN=Basic <your_base64_token>
OPENAI_API_KEY=your-openai-api-key

Instrumentation

Call OpenAIInstrumentor().instrument() before importing or instantiating the OpenAI client.

import time
from dotenv import load_dotenv
load_dotenv()

from openinference.instrumentation.openai import OpenAIInstrumentor
from openobserve import openobserve_init

OpenAIInstrumentor().instrument()
openobserve_init()

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

assistant = client.beta.assistants.create(
    name="Observability Assistant",
    instructions="Answer questions about observability concisely in one sentence.",
    model="gpt-4o-mini",
)

thread = client.beta.threads.create()
client.beta.threads.messages.create(
    thread_id=thread.id,
    role="user",
    content="What is OpenTelemetry?",
)

run = client.beta.threads.runs.create(
    thread_id=thread.id,
    assistant_id=assistant.id,
)

while run.status in ("queued", "in_progress"):
    time.sleep(0.5)
    run = client.beta.threads.runs.retrieve(thread_id=thread.id, run_id=run.id)

messages = client.beta.threads.messages.list(thread_id=thread.id)
print(messages.data[0].content[0].text.value)

client.beta.assistants.delete(assistant.id)

What Gets Captured

Each Assistants API workflow produces multiple spans — one per API call (thread creation, message creation, run creation, run retrieval, message listing).

Attribute Description
openinference_span_kind LLM
operation_name API resource called (e.g. Thread, Run)
llm_provider openai
llm_system openai
llm_observation_type GENERATION
llm_usage_tokens_input Input tokens consumed
llm_usage_tokens_output Output tokens generated
llm_usage_tokens_total Total tokens consumed
llm_invocation_parameters Request parameters sent to the API
input_mime_type application/json
output_mime_type application/json
duration Span latency
span_status OK on success, ERROR on failure

Viewing Traces

  1. Log in to OpenObserve and navigate to Traces in the left sidebar
  2. Each assistant interaction appears as a group of spans covering thread setup, run execution, and message retrieval
  3. Filter by llm_provider = openai and openinference_span_kind = LLM to isolate Assistants API spans
  4. Click any span to inspect token counts and run status

OpenAI Assistants API trace in OpenObserve

Next Steps

With the Assistants API instrumented, every thread run is recorded in OpenObserve. From here you can track token usage per assistant, monitor run latency across threads, and set alerts on failed runs.

Read More