Langfuse

Open-source LLM observability

dataintermediateFree tier

Overview

Langfuse provides tracing, evaluation, and analytics for LLM applications. Track every LLM call, measure quality with custom evals, and debug production issues. Open-source (self-host free) or cloud. Integrates with LangChain, LlamaIndex, and Vercel AI SDK.

Best For

+LLM app developers
+Production monitoring
+Quality evaluation
+Cost tracking

News & Updates

Tracium launches 1-line AI agent observability tool to track LLM costs and performance

A developer built Tracium, a simplified observability tool that tracks every AI API call to LLM providers with just one line of code. The tool was created after finding existing solutions like LangSmith and LangFuse too complex to integrate for production monitoring. Tracium provides traces, spans, and I/O visibility specifically to help teams understand agent costs and performance in real-world deployments.

NEW_TOOLS3/4/2026

XR2 Launches Prompt A/B Testing Platform to Track Business Conversions

A developer who previously ran an AI SaaS with 148,000 users has launched XR2, a prompt management platform focused on A/B testing. Unlike existing tools like Langfuse and PromptLayer that trace LLM calls, XR2 specifically tracks which prompt variants lead to user signups and purchases. This solves the core pain point of iterating on prompts without deploying new code, shifting focus from technical debugging to business outcomes.

NEW_TOOLS3/4/2026

AgentLens Launches Open-Source Observability Platform for Debugging AI Agents

A developer launched AgentLens, an open-source observability platform built specifically for debugging multi-agent AI systems. The tool provides an interactive DAG topology graph to visualize tool calls, LLM calls, and sub-agent spawns, plus time-travel replay to step through agent runs frame-by-frame. This addresses a key pain point where existing solutions like LangSmith are cloud-only and paid, while Langfuse lacks agent topology understanding.

NEW_TOOLS3/1/2026