Claude Code adds OpenTelemetry tracing for cost and tool calls
10/26/2026 — 10/26, 21:19·1 sources (all)·1 reports (all)
Key takeaways
- A tutorial shows how to enable Claude Code's built-in OpenTelemetry export: set CLAUDE_CODE_ENABLE_TELEMETRY=1, choose an exporter and point…
Background & analysis
On October 26, 2026, a post on DEV Community explained how to turn on the built-in OpenTelemetry export in Anthropic's Claude Code. The setup is straightforward: users set the environment variable CLAUDE_CODE_ENABLE_TELEMETRY=1, select an exporter, and point it at their own collector. Once enabled, Claude Code feeds token usage, cost, commit, pull request, and tool-decision data into the user's own observability stack, with data tagged by model, subagent, Skill, plugin, and MCP server.
The article also clarifies several default limits of the telemetry. Prompt text, tool details, and raw API bodies are turned off by default, and each of these has its own separate opt-in flag. For teams trying to monitor Claude Code usage across the organization, the /usage command only covers local sessions; cross-device comparisons at the team level require telemetry. There is also a deployment caveat: setting OTEL exporter variables in a project's config files has no effect, as they are ignored. As a result, rolling out the feature must be handled through managed settings instead of project-level configuration.
AI-generated from 1 reports · updated 11 hours ago
Related AI tools
- Heat index
- 73
- All sources
- 1
- All reports
- 1
- First seen
- 11 hours ago
Reports on this story headlines open the original
A walkthrough of Claude Code's built-in OpenTelemetry export: set CLAUDE_CODE_ENABLE_TELEMETRY=1, pick an exporter and point it at a collector to stream token, cost, commit, PR and tool-decision data into your own observability stack. Metrics can be split by model, subagent, skill and MCP server, while prompt text and tool details stay off by default. The piece also notes that /usage only covers the local machine.
DEV Community · AIAI score 63
Other stories people are talking about
- Heat index 71Live test finds ~10x gap in Claude Code vs Codex agent usage limits1 sources
- Heat index 51Developer maps token impact of all 33 Claude Code hook events1 sources
- Heat index 65Developer shares 50 real Claude Code errors and corresponding interception rules1 sources
- Heat index 73CabloyJS author shares Claude Code prompt for SMS suite1 sources
- Heat index 474NewAnthropic's S-1 reveals huge compute commitments, heavy customer usage of Claude5 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

