Phoenix

OverviewCopy Anchor Link

Phoenix can record calls from an OpenAI Python client whose Base URL points to TokenLab.

InstallationCopy Anchor Link

pip install arize-phoenix openinference-instrumentation-openai openai

EnvironmentCopy Anchor Link

export TOKENLAB_API_KEY="sk-your-tokenlab-key"

Register PhoenixCopy Anchor Link

Start Phoenix locally or use an existing deployment before running this example. By default, register() sends traces to the local OTLP/gRPC collector at http://localhost:4317. For another deployment, configure the collector endpoint and authentication using the Phoenix setup guide. Installing the SDK does not start the collector.

from phoenix.otel import register

tracer_provider = register(
    project_name="tokenlab-app",
    auto_instrument=True,
)

Call TokenLabCopy Anchor Link

import os

from openai import OpenAI

client = OpenAI(
    api_key=os.environ["TOKENLAB_API_KEY"],
    base_url="https://api.tokenlab.sh/v1",
)

response = client.chat.completions.create(
    model="claude-sonnet-5",
    messages=[{"role": "user", "content": "Write a short haiku."}],
)

print(response.choices[0].message.content)

Endpoint NotesCopy Anchor Link

Phoenix traces the OpenAI SDK call path. Use this page for Chat Completions style TokenLab calls. Other TokenLab native endpoints can still be traced if your application instrumentation captures the client you use for those requests.