Before you begin, make sure you have:
- A PandaProbe account. Sign up at app.pandaprobe.com.
- A PandaProbe API key + project name.
- An API key for the LLM provider you want to trace: OpenAI, Gemini, or Anthropic.
Install the SDK
pip install "pandaprobe[openai,gemini,anthropic]"
uv add "pandaprobe[openai,gemini,anthropic]"
Set environment variables
export PANDAPROBE_API_KEY="your-api-key"
export PANDAPROBE_PROJECT_NAME="your-project-name"
Self-hosting PandaProbe? Set PANDAPROBE_ENDPOINT="your-endpoint".
Wrap your LLM client
Choose the provider you want to trace. Make sure the matching provider API key is available in your environment.from pandaprobe.wrappers import wrap_openai
from openai import OpenAI
client = wrap_openai(OpenAI())
response = client.chat.completions.create(
model="gpt-5.4",
messages=[{"role": "user", "content": "What is PandaProbe?"}],
)
print(response.choices[0].message.content)
from pandaprobe.wrappers import wrap_gemini
from google import genai
client = wrap_gemini(genai.Client())
response = client.models.generate_content(
model="gemini-3.1-flash-preview",
contents="What is PandaProbe?",
)
print(response.text)
from pandaprobe.wrappers import wrap_anthropic
from anthropic import Anthropic
client = wrap_anthropic(Anthropic())
response = client.messages.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "What is PandaProbe?"}],
)
print(response.content[0].text)
View your trace
Open the PandaProbe dashboard. You should see a trace with an LLM span containing the input messages, output, model name, and token usage, all captured automatically.
What’s next?
LLM Providers
Provider wrappers for OpenAI, Anthropic, Gemini, and more
Agent Frameworks
Framework integrations for LangGraph, CrewAI, and more
Manual Instrumentation
Full control with decorators and context managers
For short-lived scripts, call pandaprobe.flush() before the process exits to ensure all traces are sent. Long-running services flush automatically.
