Skip to main content

​
Overview

This guide shows you how to integrate Latitude Telemetry into an application that uses LiteLLM to call models across providers behind a single API. Latitude wires LiteLLM’s built-in OpenTelemetry callback, so every litellm.completion call — whatever the underlying provider — is captured with messages, model, and token usage.
You’ll keep calling LiteLLM exactly as you do today. Telemetry simply observes and enriches those calls.
LiteLLM instrumentation is available in the Python SDK only.

​
Requirements

  • A Latitude account and API key
  • A Latitude project slug
  • A project that uses LiteLLM (litellm)
  • A key for whichever provider you route to (e.g. OPENAI_API_KEY)

​
Steps

1

Install

pip install latitude-telemetry litellm
2

Initialize and use

import litellm

from latitude_telemetry import Latitude, capture

latitude = Latitude(
    api_key="your-api-key",
    project="your-project-slug",
    instrumentations={"litellm": litellm},
)

def generate_reply():
    response = litellm.completion(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "Hello"}],
    )
    return response.choices[0].message.content

capture("generate-reply", generate_reply)

latitude.shutdown()

​
Seeing Your Traces

Once connected, traces appear automatically in Latitude:
  1. Open your project in the Latitude dashboard
  2. Each execution shows input/output messages, model, token usage, latency, and errors