Changelog

Set additional provider options in your LLM calls in playground and in llm-as-a-judge evaluations.

Append .md to any docs URL to fetch the page as Markdown. Built at compile time for fast, reliable access.

Day 1 support for OpenAI GPT-5 including tracking token counts and USD spend.

Assign users to annotation queues to make it easier for team members to focus on relevant tasks.

Receive prompt change notifications directly in your Slack channels with our native integration.

Get notified when your organization exceeds predefined event volume thresholds to better manage your Langfuse Cloud usage and costs.

The LLM Playground now supports side-by-side prompt comparison with parallel LLM execution.

Annotation Queues now support session-level annotation, making it easier to evaluate multi-turn interactions in your LLM applications.

Trace real-time voice AI agents and multimodal conversations built with LiveKit Agents via OpenTelemetry and Langfuse

Enable non-technical users to trigger external experiment pipelines directly from the Langfuse UI on datasets.

Find prompts by searching through their actual content with our new full-text search capability

Self-hosted deployments can now use the AWS SDK default credential provider chain for Amazon Bedrock LLM connections.

Get real-time notifications for new prompt versions, changes to deployment labels or tags, and other edits.

Use Langfuse's prompt management capabilities directly in your n8n workflows to build powerful no-code agents.

Langfuse says hello to Gemini 2.5 Pro (GA), Gemini 2.5 Flash (GA), and Gemini 2.5 Flash Lite (Preview). Support added for the LLM playground, evaluations, and cost tracking.

Inject context or message history into your chat prompts at just the right spot.

Analyze multi-dimensional data with the new pivot table widget in Langfuse custom dashboards.

Now it is easier to find prompts in prompt management via the new search bar

Visualize data distributions with the new histogram chart type in Langfuse custom dashboards.

Integrate Langfuse into your existing codebase using AI agents like Cursor, Claude Code, GitHub Copilot, and Windsurf. Powered by the new Langfuse Docs MCP server.

Simplified JSON handling in prompts designated for LangChain consumption - no more extra escaping required.

Datasets and Audit Logs can now be exported as CSV and JSON directly from the UI.

Download metrics from any chart in your custom dashboards as CSV files for further analysis and reporting.

Use slashes in prompt names to create folders for better organization.

Langfuse adds same day support for OpenAI's o3-pro model and updated o3 pricing

The OpenTelemetry-based Python SDK v3 is now stable and ready for production use.

We have Open Sourced all Remaining Product Features under the MIT License

Trace realtime audio agents built with Pipecat via OpenTelemetry and Langfuse

Introducing the Langfuse Evaluator Library with prebuilt evaluators. Plus, enjoy a revamped UX with trace and variable previews for easier LLM evaluation.

The latest generation of the Langfuse Python SDK, built on the OpenTelemetry framework.

Langfuse says hello to Anthropic Claude Sonnet 4 & Opus 4 with day 1 support for both the LLM playground, LLM-as-a-judge and cost tracking.

Deploy Langfuse on AWS, GCP, or Azure using our new Terraform modules.

Create your own custom dashboards or use Langfuse-curated dashboards to analyze and understand your AI application deeply.

Pin your favorite table settings and share them with a single link.

Find content in inputs and outputs of your traces and observations with full text search.

Control which product modules are visible in the main menu through environment variables when self-hosting Langfuse

Build custom reports and dashboards with our flexible Metrics API. Query traces, observations, and scores with custom dimensions, metrics, and time granularity.

Score dataset runs to assess the overall quality of each run

Create and manage scores at the session level for more comprehensive evaluation of conversational AI applications

Easily copy the source markdown of documentation pages with a single click. Paste it into ChatGPT, Cursor, Windsurf and move way faster while using Langfuse.

Langfuse launches same day full compatibility with OpenAI's latest models o3 and o4-mini

Langfuse now offers compatibility for OpenAI's new series of GPT models 4.1, 4.1-mini and 4.1-nano.

Langfuse now supports administrative actions via APIs across organizations and projects.

Langfuse now supports protected prompt labels for better access control over your prompt deployment workflow

Langfuse says hello to Gemini 2.5 Pro Experimental. Support added for the LLM playground.

Define and test tool calls and structured output schemas directly in the Langfuse Playground

Quickly navigate trace details in the table view via keyboard shortcuts or by clicking on the table rows.

Our redesigned prompt view brings enhanced visualization, version comparison, and search capabilities.

Introducing our redesigned trace view with improved hierarchy visualization, timeline/tree toggle, and powerful search capabilities.

Trace your OpenAI Response API calls with our Python and JS SDKs