Tools · SPBoucher
Menu

Langfuse

In-depth entry

by Langfuse (part of ClickHouse, Inc. since January 2026) · Platform

Open-source LLM observability and evaluation platform: tracing, LLM-as-judge and human evals, prompt management, datasets and cost dashboards, self-hostable or as Langfuse Cloud

MIT licenseOpen sourceFreemiumPaidEnterpriseSelf-hostedCloudSelf-hostable
Visit official site Compare

Link verified Nov 2, 2026

Overview

Langfuse is an open-source platform for observing and evaluating LLM applications and agents. Instrumented applications send hierarchical traces of model calls, tool invocations and retrieval steps, which can be filtered by user, session, cost, latency or metadata. On top of traces sit evaluations (LLM-as-a-judge, heuristic scorers and human annotation queues), versioned prompt management with deployments and rollbacks, datasets and experiments for regression testing, and dashboards with cost and latency alerts. Native SDKs exist for Python and TypeScript, the platform is OpenTelemetry-native so other languages instrument through OTel, and more than a hundred integrations cover frameworks (LangChain, LlamaIndex, CrewAI, OpenAI Agents, Pydantic AI, LiteLLM), model providers and coding tools such as Claude Code and Cursor.

All product features are MIT-licensed; the repository notes a small ee/ directory under a separate license, and the company states there are no feature gates on self-hosting. Deployment is Docker Compose, Kubernetes via Helm or Terraform templates for AWS, GCP and Azure, with ClickHouse as the analytical store. Langfuse Cloud offers Hobby (free), Core, Pro and Enterprise plans with US and EU regions. In January 2026 Langfuse GmbH was acquired by ClickHouse, Inc.; the team stated that the roadmap, open-source license and self-hosting commitment were unchanged.

For a team running agents on its own infrastructure, Langfuse is one of the few observability stacks that can be fully self-hosted without licensing fees. Operating it does mean running Postgres, ClickHouse, Redis and object storage, so very small projects may prefer the hosted tier.

Capabilities

  • Hierarchical tracing of LLM, tool and retrieval calls with sessions and users
  • LLM-as-a-judge, heuristic and human-annotation evaluations
  • Versioned prompt management with deployments, rollbacks and a playground
  • Datasets and experiments for regression testing
  • Cost and latency dashboards with alerts
  • Python and TypeScript SDKs; OpenTelemetry-native ingestion
  • 100+ integrations including LangChain, LlamaIndex, LiteLLM, OpenAI Agents and Claude Code
  • Self-hosting via Docker Compose, Helm or Terraform; Langfuse Cloud in US and EU

Typical workflows

  1. Instrument an agent and inspect failing traces by session
  2. Run nightly LLM-as-judge evaluations on sampled production traffic
  3. Manage and roll back production prompts without redeploying code
  4. Self-host on a Kubernetes cluster with ClickHouse and Postgres

Strengths

  • MIT-licensed with no self-hosting feature gates
  • Framework-agnostic through OpenTelemetry
  • Mature prompt management alongside tracing
  • Hosted and self-hosted options with EU region

Limitations

  • Self-hosting requires Postgres, ClickHouse, Redis and object storage
  • Hosted plans meter ingested units and retention
  • Corporate ownership changed in 2026; long-term direction to be watched
  • Enterprise self-hosted support is a paid add-on

Prerequisites

  • Application instrumented with an SDK or OpenTelemetry
  • Docker or Kubernetes for self-hosting, or a Langfuse Cloud account

Who it is for

  • Teams operating LLM applications in production
  • Researchers tracking experiments on prompts and models
  • Organizations requiring self-hosted observability

Alternatives & comparison

LangSmith is LangChain's hosted equivalent, with tighter LangGraph integration and self-hosting limited to the Enterprise plan. Braintrust focuses on evals and agent trace analysis as a hosted service with on-prem only at enterprise scale. Weights & Biases Weave covers LLM tracing inside a broader ML-experiment platform. Promptfoo is a local-first CLI for evals and red teaming rather than a tracing backend; it is often used together with Langfuse.

  • Braintrust

    Hosted observability and evaluation platform for AI agents with tracing, experiments, pattern detection, a CLI, MCP access and an AI assistant (Loop) that analyzes traces and builds evals

    MixedFreemium
  • LangSmith

    LangChain's hosted platform for tracing, evaluating and deploying LLM applications and agents, with prompt tooling, an LLM gateway and cloud, hybrid or self-hosted options

    ProprietaryFreemium
  • LiteLLM

    Open-source Python SDK and self-hosted AI gateway that calls 100+ LLM providers in the OpenAI format, with routing, fallbacks, virtual keys, budgets, spend tracking and logging callbacks

    MixedOpen source
  • promptfoo

    Open-source, local-first CLI and library for evaluating prompts, models, RAG pipelines and agents, with automated red teaming and CI integration

    MixedOpen source
  • Ragas

    Open-source Python library for evaluating LLM and RAG applications with LLM-driven and custom metrics, experiments, dataset management and result tracking

    Open sourceOpen source
  • Weights & Biases

    ML experiment tracking, sweeps, artifacts and model registry plus Weave for LLM tracing and evaluation, now sold by CoreWeave within its Forge platform with a free academic tier

    MixedFreemium
Compare Langfuse with its alternatives →

Sources & verification

Primary sources consulted for this record. Claims are tied to the fields each source supports; the automated link checker re-verifies every URL on a schedule.

Confidence
High
Reviewed
Nov 1, 2026
Links
Verified · Nov 2, 2026

Spotted an error?

Broken link, outdated license, wrong category, or a tool that no longer exists? Tell us; reports are reviewed by the editor.

LiteLLM

BerriAI · Prompting, evaluation & observability

In-depth entry

Open-source Python SDK and self-hosted AI gateway that calls 100+ LLM providers in the OpenAI format, with routing, fallbacks, virtual keys, budgets, spend tracking and logging callbacks

MITOpen sourceFreeLocalSelf-hostedNov 2, 2026

OpenRouter

OpenRouter, Inc. · Model providers & APIs

Unified API gateway to hundreds of models from many providers with automatic fallbacks, credit-based pass-through billing, BYOK, free models and data-policy routing

ProprietaryFreemiumUsage-basedBroken

DSPy

Stanford NLP and the DSPy community · Prompting, evaluation & observability

Python framework for programming language-model pipelines with typed signatures and modules, then optimizing prompts and weights automatically against a metric

MITOpen sourceFreeLocalSelf-hostedNov 2, 2026

Ragas

Vibrant Labs (formerly Exploding Gradients) · Prompting, evaluation & observability

Open-source Python library for evaluating LLM and RAG applications with LLM-driven and custom metrics, experiments, dataset management and result tracking

Apache-2.0Open sourceFreeLocalBroken