View a trace

Inspect a trace in LangSmith using the Messages view or Details view.

When you open a trace in the LangSmith UI, two views are available at the top of the trace panel:

• Messages: Renders a trace as a readable conversation, showing the sequence of messages exchanged between the user, assistant, and any tools called.
• Details: Shows the complete technical record of a trace, including every run with its inputs, outputs, timing, token counts, and metadata.

Messages view

The Messages view is the default view for traces under 20 MB.

LangSmith does not display the Messages view for traces larger than 20 MB. On self-hosted deployments, this limit is configurable via environment variable.

Use the Messages view to understand what happened in an agent interaction quickly: what the user asked, how the assistant responded, and which tools the agent invoked.

Customize the Messages view

You can control how runs appear in the Messages view using metadata keys on individual runs.

• ls_agent_type: Accepts agent, subagent, or middleware. Used to filter middleware runs from the Messages view and to control how subagent details are displayed.

• ls_message_view_exclude: Set to true to exclude a run and all runs in its subtree from the Messages view.

  @traceable(metadata={"ls_message_view_exclude": True})
  def my_run():
      ...
  

Provider-specific behavior

The messages shown in the Messages view depend on the integration:

Integration	Messages shown
Anthropic / Claude SDK	Messages from the current trace only
DeepAgents / OpenAI	Messages from the current trace and ancestor traces

Details view

Use the Details view to debug execution: inspect a specific LLM call, review tool inputs and outputs, or identify where an error occurred.

Customize the Details view

Setting run_type="llm" on a run causes the Details view to render token counts and latency for that run. See Log an LLM trace for the full message format specification.

Actions

From the Details view, you can also:

• Compare traces: Select two traces to open a side-by-side comparison. See Manage a trace.
• Share a trace: Generate a public link to the trace. See Manage a trace.
• View server logs: Access server logs associated with a trace generated by a LangSmith deployment. See Manage a trace.

---

Edit this page on GitHub or file an issue.

Connect these docs to Claude, VSCode, and more via MCP for real-time answers.