Automatic Model Call Logging

Use the Planck Async Logger to observe existing model calls without routing them through the Planck AI Gateway. Your application continues calling OpenAI, Anthropic, or another supported provider directly, and the logger sends trace data to Planck in the background.

This is a Planck integration. Install @inquantum/planck-async for Node.js or planck-async for Python. No additional logging service or account is required.

Set up automatic logging

Node.js
Python

Install the Planck Async Logger

npm install @inquantum/planck-async

Initialize Logger

import { PlanckAsyncLogger } from "@inquantum/planck-async";
import OpenAI from "openai";

const logger = new PlanckAsyncLogger({
  apiKey: process.env.PLANCK_API_KEY,
  // pass in the providers you want logged
  providers: {
    openAI: OpenAI,
    //anthropic: Anthropic,
    //cohere: Cohere
    // ...
  }
});
logger.init();

const openai = new OpenAI();

async function main() {
  const completion = await openai.chat.completions.create({
    messages: [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Who won the world series in 2020?"},
      {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
      {"role": "user", "content": "Where was it played?"}
    ],
    model: "gpt-4o-mini",
  });

  console.log(completion.choices[0]);
}

main();

Properties

You can set properties on the logger to be used in Planck using the withProperties method. (These can be used for Sessions, User Metrics, and more.)

const sessionId = randomUUID();

logger.withProperties({
  "Planck-Session-Id": sessionId,
  "Planck-Session-Path": "/abstract",
  "Planck-Session-Name": "Course Plan",
}, () => {
  const completion = await openai.chat.completions.create({
    // ...
  })
})

Install the Planck Async Logger

pip install planck-async

Initialize Logger

from planck_async import PlanckAsyncLogger
from openai import OpenAI

logger = PlanckAsyncLogger(
  api_key=PLANCK_API_KEY,
)

logger.init()

client = OpenAI(api_key=OPENAI_API_KEY)

# Make the OpenAI call
response = client.chat.completions.create(
  model="gpt-4o-mini",
  messages=[
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Who won the world series in 2020?"},
    {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
    {"role": "user", "content": "Where was it played?"}
  ]
)

print(response.choices[0])

Properties

You can set properties on the logger to be used in Planck using the set_properties method. (These can be used for Sessions, User Metrics, and more.)

session_id = str(uuid.uuid4())

logger.set_properties({
  "Planck-Session-Id": session_id,
  "Planck-Session-Path": "/abstract",
  "Planck-Session-Name": "Course Plan",
})

response = client.chat.completions.create(
  # ...
)

Disable logging

You can completely disable all logging to Planck if needed when using the async integration mode. This is useful for development environments or when you want to temporarily stop sending data to Planck without changing your code structure.

Python
Node.js

# Disable all logging in async mode
logger.disable_logging()

# Later, re-enable logging if needed
logger.enable_logging()

When logging is disabled, no traces will be sent to Planck. This is different from disable_content_tracing() which only omits request and response content but still sends other metrics. Note that this feature is only available when using Planck’s async integration mode.

Supported providers

Other integration options

Gateway Integration

Get Started

Models & Inference

Reliability & Controls

Observability

Prompt Management

Integrations

Enterprise

Automatic Model Call Logging

Set up automatic logging

Disable logging

Supported providers

Other integration options

​Set up automatic logging

​Disable logging

​Supported providers

​Other integration options

Set up automatic logging

Disable logging

Supported providers

Other integration options