Integration

Traceloop OpenLLMetry with Meridian

Traceloop's OpenLLMetry SDK auto-instruments the official OpenAI client. Because Meridian uses the same client, you get full OTLP traces with zero extra wiring — just init Traceloop before constructing the client.

Install

pip install traceloop-sdk

Init + call

from traceloop.sdk import Traceloop
from openai import OpenAI

Traceloop.init(app_name="my-app", api_key="tl_***")

client = OpenAI(
    api_key="nim_live_***",
    base_url="https://meridian.getnimbus.net/api/v1",
)

# Every call is auto-traced. Pretrip the spans without any extra code.
resp = client.chat.completions.create(
    model="azure-swc/gpt-4.1",
    messages=[{"role": "user", "content": "ping"}],
)

What gets traced

  • Request prompt + parameters
  • Response content + usage (tokens, finish reason)
  • Latency p50/p95 + region routing
  • Tool calls + function arguments
← All integrationsObservability options