Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.openlit.io/llms.txt

Use this file to discover all available pages before exploring further.

The OpenLIT platform provides comprehensive trace visualization capabilities. You can view traces in two ways:
  1. Traces Page: Navigate to the Traces section at 127.0.0.1:3000/traces to view all distributed traces from your AI applications with detailed span analysis and execution flow.
  2. Dashboard Widgets: Create custom trace widgets in your dashboards to monitor specific trace metrics, latency trends, and performance insights alongside other observability data.

​
AI analysis for traces and spans

Trace details include an AI Analysis tab for both the full trace hierarchy and individual spans. Use it to turn raw OpenTelemetry data into a structured review of performance, reliability, cost, token usage, and execution flow. The analysis works at two scopes:
  • Trace hierarchy analysis: Reviews the root span and child spans together so you can understand the full request, agent path, tool calls, model calls, retries, errors, and cost drivers.
  • Individual span analysis: Focuses on one selected span when you need to inspect a specific model call, tool execution, retrieval step, error, or latency hotspot.
Each analysis run evaluates:
  • Performance: latency, slow spans, blocking operations, duration hotspots, and inefficient paths.
  • Reliability: errors, failed spans, exception signals, retries, and unstable dependencies.
  • Cost: model usage, estimated spend, high-cost spans, and cost optimization opportunities.
  • Token efficiency: input tokens, output tokens, cache usage, repeated context, and verbose responses.
  • Prompt and model behavior: prompt structure, model choice, output quality, and response efficiency.
  • Telemetry quality: missing attributes, incomplete trace context, inconsistent service metadata, and weak observability signals.
  • Actionability: a grader pass reviews each section so recommendations are clearer and easier to apply.

​
Run analysis from trace details

  1. Open a trace from the Traces page or from a trace widget.
  2. Select AI Analysis in the trace detail view.
  3. Choose whether to analyze the full trace hierarchy or the currently selected span.
  4. Review the streamed progress. When the run completes, the analysis collapses into a reusable result.
Existing runs are shown in the analysis panel, so you can revisit prior findings without rerunning the model. Use rerun only when the trace data or investigation question has changed.

​
Run analysis from Otter

You can also ask Otter to analyze traces or spans by natural language: 
"Analyze this trace hierarchy"
"Review span <span_id> for token waste"
"Find cost optimization opportunities for traces where session.id is dc7b1ba1-..."
"Analyze the slowest traces from the last 24 hours"
When Otter references a trace or span in its answer, OpenLIT renders clickable trace/span pills so you can jump back into the trace detail experience.

​
Group traces

Use Group By on the Traces page to roll up large trace lists into meaningful groups before drilling into individual spans. Grouping works with the selected time range and any active filters, so you can narrow the dataset first and then compare trace segments. You can group traces by:
  • Model: Compare requests by gen_ai.request.model.
  • Provider: Compare requests by gen_ai.system.
  • Span Name: Group repeated operations or framework steps.
  • Application: Compare services using the service.name resource attribute.
  • Custom attribute: Group by any span attribute, resource attribute, or top-level trace field available in your trace data.
Grouped rows show the number of spans in each group, total cost, token usage, and average duration. Click a group row to drill into the matching traces. The breadcrumb above the table shows the current grouping path and lets you return to all groups or remove grouping.
Grouping is best for finding high-volume models, expensive providers, slow span types, or application-level hotspots before opening an individual trace.

​
Filter and group together

Grouping can be combined with the existing trace filters. For example, you can filter to a single environment, apply a maximum cost threshold, and then group by model to find which models dominate that filtered slice. Custom attribute filters and custom group-by attributes can be used together when you need to inspect application-specific metadata.

Quickstart: LLM Observability

Production-ready AI monitoring setup in 2 simple steps with zero code changes

Create a dashboard

Create custom visualizations with flexible widgets, queries, and real-time AI monitoring

Integrations

60+ AI integrations with automatic instrumentation and performance tracking

Zero-code observability with the OpenLIT Controller

Discover and instrument LLM traffic across Kubernetes, Docker, and Linux using eBPF — no code changes required.