Langfuse

https://langfuse.com/

Memory module reads/writes captured as named spans. Trace Log View concatenates every agent step including memory ops. Agent Graphs (GA 2025) infer graph structure from observation nesting; session-level replay tracks how memory state evolves.

At a glance

Type
Memory ops as named spans + Agent Graphs
Tier
T1
Created
2023-03 (Langfuse founded March 2023; LLM observability beta summer 2023; GA Feb 2024)
Latest release
not applicable — not OSS
License
not applicable — not OSS
GitHub
not applicable — no GitHub repo
Pricing
free: 50K observations/mo; Pro: $59/mo; Team: $199/mo; Enterprise: custom; acquired by ClickHouse so pricing evolut…
Funding
acquired · 2026-01

Taxonomy

storage
relational
retrieval
exact-match
persistence
long-term
update
append-only
unit
episode
governance
auditable
conflict
n/a

When to use

Optimised for: memory operation tracing + drift / poisoning detection

Anti-fit: not for use cases that don't run agent workloads in production

Pros & cons

Pros

OSS-first LLM observability with a polished cloud option — self-hostable for compliance-sensitive shops; large open-source community.

Cons

Memory-specific observability is improving but less domain-specialized than Galileo's purpose-built memory views.

Claims & capabilities

No automated drift / poisoning detection — structural observability only. Free 50k observations; Pro $59/mo; Team $199/mo.

Technical surface

API surface
REST, SDK: Python, JS/TS
Backend storage
Postgres + ClickHouse
Deployment
SaaS cloud (Berlin + SF) or self-hosted Docker/K8s; ClickHouse native integration post-acquisition
Embedding model
not applicable — observability product
Multi-tenancy
namespace (org/project)
MCP
via official adapter — langfuse-mcp
A2A
no data — searched langfuse.com, langfuse.com/docs, langfuse.com/integrations; A2A protocol not advertised (extensive framework integrations, MCP server present, OTel-native)
OpenTelemetry
first-class — native OTel SDK

Compare Langfuse with…

Similar systems

Other memory observability & monitoring in the catalog, ranked by inbound references.

  • AgentOps T2

    When Mem0 is connected, gains Memory Operation Timeline, Search Analytics, Memory Growth tracking, Error Tracking per memory call. Standalone, records context at each step but doesn't analyse memory quality.

  • Galileo (galileo.ai) T1

    Treats memory as first-class in multi-agent tracing. Luna-2 SLMs (3B / 8B) scan every interaction for intent drift and belief drift; 20+ checks at sub-200ms latency. Catches when agent A's view of the world splits from teammate B's. OpenTelemetry-compatible.

  • LangSmith T1

    Memory reads, vector DB retrievals, state changes are distinct span types in traces. RAG eval separates retrieval quality (context precision) from generation quality (faithfulness). Dataset versioning guards against eval drift.

  • Ratine T2

    Only tool found that scans the persistent memory layer on disk rather than runtime tracing. Detects injected instructions, obfuscated payloads (zero-width Unicode, base64, homoglyphs, hex), C2-pattern URLs, credential leakage. ratine diff compares snapshots for belief drift.

Row last verified 2026-05-14. Catalog data is CC-BY-4.0 — see how to read this.