OpenAI Codex OpenTelemetry

• Available

  Metrics

  Quantitative telemetry covering supported usage, token and execution measurements, exported through the otel.metrics_exporter setting.

• Available

  Events

  Structured events describing supported Codex activity including model requests, user interactions, tool execution and approval decisions.

• Available

  Traces

  Distributed traces providing visibility into operations across supported Codex agent workflows, configured with otel.trace_exporter.

Telemetry availability can differ between Codex execution modes and versions. Always refer to the current Codex documentation for mode-specific support.

Common attributes

• otel.environment

  Environment tag applied to emitted OpenTelemetry events. Default: dev.

• service.name

  Service identity of the Codex client emitting telemetry, for example codex-cli.

• otel.span_attributes

  Static attributes applied to exported trace spans.

• otel.tracestate

  W3C tracestate members propagated with exported trace context.

• Adoption & Usage

  Understand how Codex usage evolves across engineering workflows by analyzing supported sessions and agent activity over time.

• AI Consumption

  Analyze model requests and token consumption to better understand how AI resources are being used.

• Tools & Agent Workflows

  Observe how Codex invokes tools, handles approvals and executes tasks across agentic development workflows.

• Performance & Reliability

  Use supported timing, request and execution telemetry to investigate agent performance and operational issues.

  Usage and activity metrics provide context about AI-assisted development. They should not be treated as standalone measures of developer productivity.

# ~/.codex/config.toml
[otel]
environment = "production"
log_user_prompt = false

# Logs and events exporter
exporter = "otlp-http"

[otel.exporter.otlp-http]
endpoint = "https://otlp.example.com"
protocol = "binary"
headers = { "Authorization" = "Bearer ${OTLP_TOKEN}" }

# Traces exporter
trace_exporter = "otlp-http"

[otel.trace_exporter.otlp-http]
endpoint = "https://otlp.example.com"
protocol = "binary"

# Metrics exporter
metrics_exporter = "otlp-http"

Keep telemetry disabled or restrict exported content when observability is not required.

• User Prompts

  Prompt text should not be treated as required telemetry. Codex provides the otel.log_user_prompt setting, an explicit opt-in for exporting raw user prompts with OpenTelemetry logs.

• Tool Activity

  Tool execution telemetry can contain information about development workflows and should be handled according to organizational security and privacy policies.

• Telemetry Destination

  OpenTelemetry export allows organizations to send supported Codex telemetry to their chosen compatible infrastructure, including TLS settings and static headers per exporter.

Explore Other OpenTelemetry Integrations

• 

  Claude Code

  Claude Code OpenTelemetry: metrics, events and traces, configuration and privacy considerations.

  Read the guide
• 

  Gemini CLI

  Gemini CLI OpenTelemetry: available metrics, logs and traces, setup and privacy considerations.

  Read the guide
• 

  GitHub Copilot

  GitHub Copilot OpenTelemetry: agent metrics, events and traces, setup and privacy considerations.

  Read the guide
• 

  Cursor

  Cursor OpenTelemetry: exported metrics and logs, destination setup and privacy considerations.

  Read the guide
• 

  Qwen Code

  Qwen Code OpenTelemetry: metrics, logs and traces, OTLP setup and privacy considerations.

  Read the guide