Histogram Charts in Custom Dashboards
Steffen Schmitz
Visualize data distributions with the new histogram chart type in Langfuse custom dashboards.
We’re excited to introduce histogram charts to Langfuse custom dashboards! This new visualization option helps you understand the distribution of your data, making it easier to identify patterns, outliers, and trends in your LLM application metrics.
What are Histogram Charts?
Histogram charts display the frequency distribution of continuous data by grouping values into bins. Unlike bar charts that show discrete categories, histograms reveal the shape and spread of your data, helping you answer questions like:
- Latency Analysis: How are response times distributed across your application?
- Cost Patterns: What’s the typical range of costs per request?
- Score Distributions: How are quality scores spread across different ranges?
- Token Usage: What’s the distribution of input/output token consumption?
Getting Started
Creating a histogram chart is simple:
- Navigate to Dashboards → Widgets → New Widget
- Select your data source (traces, observations, or scores)
- Choose Histogram as your chart type
- Select the metric you want to visualize (e.g., latency, cost, tokens)
- Apply any filters to focus on specific data subsets
- Save and add to your dashboard
Learn More
For detailed instructions on creating and customizing histogram charts, visit our Custom Dashboards documentation.
Explore how histogram charts can enhance your LLM application monitoring and help you make data-driven optimization decisions.
Questions or Feedback?
We’d love to hear how you’re using histogram charts in your dashboards! Share your feedback and use cases in our GitHub Discussions.
Please open a GitHub issue if you encounter any problems or have suggestions for improvements.