Otelite Dashboard

# Set these environment variables before running Claude Code.
# Telemetry is off by default — CLAUDE_CODE_ENABLE_TELEMETRY=1 is required.
export CLAUDE_CODE_ENABLE_TELEMETRY=1
export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4318
export OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf
export OTEL_SERVICE_NAME=claude-code
export OTEL_LOGS_EXPORTER=otlp
export OTEL_METRICS_EXPORTER=otlp

# Optional: distributed traces (beta) — interaction / llm_request / tool spans
export CLAUDE_CODE_ENHANCED_TELEMETRY_BETA=1
export OTEL_TRACES_EXPORTER=otlp

# Optional: include prompt / tool input content in events (off by default)
# export OTEL_LOG_USER_PROMPTS=1
# export OTEL_LOG_TOOL_DETAILS=1

Official docs: docs.claude.com › Claude Code › Monitoring usage — authoritative list of env vars, metric names, event names, and attributes.

# The Agent SDK runs the Claude Code CLI as a child process,
# so the same env vars apply. Set them before launching your agent:
export CLAUDE_CODE_ENABLE_TELEMETRY=1
export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4318
export OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf
export OTEL_LOGS_EXPORTER=otlp
export OTEL_METRICS_EXPORTER=otlp
# Note: do NOT use OTEL_*_EXPORTER=console under the SDK
# (console output collides with the SDK's message channel).

Official docs: code.claude.com › Agent SDK › Observability with OpenTelemetry.

from opentelemetry import trace
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor

exporter = OTLPSpanExporter(endpoint="http://localhost:4318/v1/traces")
provider = TracerProvider()
provider.add_span_processor(BatchSpanProcessor(exporter))
trace.set_tracer_provider(provider)

tracer = trace.get_tracer(__name__)
with tracer.start_as_current_span("my-operation"):
    print("Hello from Python!")

Official docs: opentelemetry.io › Python.

uv pip install "mellea[telemetry]"
export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4317
export MELLEA_TRACE_APPLICATION=true
export MELLEA_TRACE_BACKEND=true
export MELLEA_METRICS_ENABLED=true
export MELLEA_METRICS_OTLP=true

Mellea is a Python library for writing generative programs — structured outputs, verifiable requirements, automatic retries. Backend spans follow OTel GenAI semconv so the Usage page and trace waterfall work out of the box. See mellea.ai · telemetry docs · github.com/generative-computing/mellea.

# pip install opentelemetry-instrumentation-openai
from opentelemetry.instrumentation.openai import OpenAIInstrumentor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
import openai

exporter = OTLPSpanExporter(endpoint="http://localhost:4318/v1/traces")
provider = TracerProvider()
provider.add_span_processor(BatchSpanProcessor(exporter))
OpenAIInstrumentor().instrument()

client = openai.OpenAI()
response = client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Hello!"}]
)
# Default: prompts/completions are captured on span attributes.
# Disable with: export TRACELOOP_TRACE_CONTENT=false

Official docs: traceloop/openllmetry › opentelemetry-instrumentation-openai.

# pip install opentelemetry-instrumentation-anthropic
from opentelemetry.instrumentation.anthropic import AnthropicInstrumentor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
import anthropic

exporter = OTLPSpanExporter(endpoint="http://localhost:4318/v1/traces")
provider = TracerProvider()
provider.add_span_processor(BatchSpanProcessor(exporter))
AnthropicInstrumentor().instrument()

client = anthropic.Anthropic()
message = client.messages.create(
    model="claude-sonnet-4-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}],
)

Official docs: traceloop/openllmetry › opentelemetry-instrumentation-anthropic. OpenLLMetry also covers Bedrock, Cohere, Gemini, Groq, Mistral, Ollama, VertexAI — same pattern; see the full list.

# pip install openinference-instrumentation-openai opentelemetry-sdk \
#             opentelemetry-exporter-otlp
from openinference.instrumentation.openai import OpenAIInstrumentor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
import openai

provider = TracerProvider()
provider.add_span_processor(BatchSpanProcessor(
    OTLPSpanExporter(endpoint="http://localhost:4318/v1/traces")))
OpenAIInstrumentor().instrument(tracer_provider=provider)

# Swap OpenAIInstrumentor for AnthropicInstrumentor (from
# openinference.instrumentation.anthropic) for Claude apps, or for
# LangChainInstrumentor, LlamaIndexInstrumentor, etc.

Official docs: Arize-ai/openinference — covers OpenAI, Anthropic, Bedrock, Mistral, LangChain, LlamaIndex, DSPy, CrewAI, LiteLLM, and more.

const { NodeTracerProvider } = require('@opentelemetry/sdk-trace-node');
const { OTLPTraceExporter } = require('@opentelemetry/exporter-trace-otlp-http');
const { BatchSpanProcessor } = require('@opentelemetry/sdk-trace-base');

const provider = new NodeTracerProvider();
provider.addSpanProcessor(new BatchSpanProcessor(new OTLPTraceExporter({
  url: 'http://localhost:4318/v1/traces'
})));
provider.register();

exporter, _ := otlptracehttp.New(context.Background(),
    otlptracehttp.WithEndpoint("localhost:4318"),
    otlptracehttp.WithInsecure(),
)
provider := trace.NewTracerProvider(trace.WithBatcher(exporter))
otel.SetTracerProvider(provider)

# Install: https://github.com/equinix-labs/otel-cli
otel-cli exec --endpoint http://localhost:4318 --service my-test -- echo "Hello Otelite!"