LangSmith unifies tracing across coding agents
Start here
LangChain's LangSmith now traces sessions from Claude Code, Codex, Cursor, GitHub Copilot Chat, Pi and OpenCode into a single trace model, using the same metadata and query syntax across tools. Token usage, cost per session, tool calls and subagent activity are normalized across these tools, and LangSmith Engine flags wasteful patterns while LLM Gateway caps spend at the user, team and organization level, LangChain says. Setup differs by tool, with separate steps for each. LangChain notes that organizations fully standardized on one tool whose native dashboard already answers their questions may not need this second layer yet, and that teams do not need all four components on day one; it recommends observability as the starting point for teams in early adoption.
Generated from reports · updated 11 h ago
Report timeline
Follow the reports to see the event from different sides.
- LangChain BlogLangSmith adds unified tracing for Claude Code, Codex, Cursor and Copilot to fix coding agent costs
LangSmith now traces sessions from Claude Code, Codex, Cursor, GitHub Copilot Chat, Pi and OpenCode into one trace model with the same metadata and query syntax. Token usage, cost per session, tool calls and subagent activity are normalized across tools, while Engine flags wasteful patterns and LLM Gateway caps spend at user, team and org level.
Heat for this event
- Current, comparable range
- 5
- Peak, comparable range
- 7November 19 17:00
- Change in 24 hours
- –
The trend compares the same participants over hours observed continuously, so its range can be narrower than the current heat count. Hover or tap the chart for hourly heat; the arrow keys move through it.