v0.21.0A security release: what an agent or a client sends Iris is bounded before Iris reads it, and a tool argument can narrow what the operator allows but never widen it→

Comparison · Observability

Iris vs Langfuse

MCP-Native Agent Eval vs SDK-Based Tracing.

TL;DR

Iris is an MCP server your agent discovers and uses on connect — one config block, no SDK, one SQLite file, every rule's precision and recall published at iris-eval.com/proof. Langfuse — from its own pages, read 2026-10-01: Langfuse is an open-source AI engineering platform with native Python/JS SDKs, 100+ library/framework integrations, prompt management, and enterprise compliance (SOC 2 Type II, ISO 27001, HIPAA). If you need broad framework support, prompt versioning, or enterprise certifications today, Langfuse is the more established choice. (source)

Iris grades what an agent did with its tools — the trace, the answer, the cost — not whether an MCP server honours its own contract; a server test harness answers that question, and Iris runs beside it. For the method, see the agent eval guide.

Feature comparison

Side by side.

13 features, the same 13 on every comparison. Every Langfuse cell links the page it was read from and the date; where the sentence it was read from was found verbatim on a plain download of that page (checked 2026-09-22), the link carries it as its title, and where it was not, the cell says so. The highlighted cells are Iris's own call on which side is stronger for a team running MCP agents — 3 to Iris, 3 to Langfuse — not a measurement.

FeatureIrisLangfuse
Integration methodOne block in the MCP config, no code — the agent discovers Iris and its tools on connectSDK imports + @observe decoratorsLangfuse's page · read 2026-10-01
Self-hostingOne process, one SQLite file; Docker image with a health checkPostgreSQL + ClickHouse + Redis + S3 + 2 containersLangfuse's page · read 2026-10-01
Where it runsNothing in the agent's process — Iris is a separate server the agent callsAsync SDK tracing in your processLangfuse's page · read 2026-10-01
Evaluation26 built-in deterministic rules and 9 custom-rule types, in-process; 7 judge templates on a key you supply; every rule's precision and recall publishedLLM-as-a-Judge, Jev as a judge (a decision model), human annotation, custom scores via API/SDK, and server-side code evaluators (Python or TypeScript)Langfuse's page · read 2026-10-01
Cost trackingPer-trace USD cost and tokens; a cost spike judged against the agent's own historyUsage and USD cost per LLM call; dashboards and Metrics API by model, tags, or userLangfuse's page · read 2026-10-01
MCP supportProtocol-native — Iris is an MCP server with 12 tools; OTLP traces inMCP server in Cloud and self-hosted Langfuse: read and write tools for prompts, traces, scores, evaluators, datasets, experiments, dashboardsLangfuse's page · read 2026-10-01
LicenseMIT, the whole packageMIT core + commercial enterprise modulesLangfuse's page · read 2026-10-01
OwnershipIndependent and founder-ledAcquired by ClickHouse (Jan 2026)Langfuse's page · read 2026-10-01
DashboardA local dashboard on its own port: traces, moments, regressions, five viewsCustomizable multi-dimension dashboardsLangfuse's page · read 2026-10-01
Framework supportAny MCP client (2 verified, 8 claimed — see /clients); OTLP/HTTP from anything else100+ library/framework integrations, plus native Python/JS SDKs and OpenTelemetryLangfuse's page · read 2026-10-01
Prompt managementNot includedVersioned prompts with labels for environment deployment; server- and client-side prompt cachingLangfuse's page · read 2026-10-01
Enterprise and complianceSelf-hosted. Nothing leaves your machine unless you turn one of these on: an OpenTelemetry endpoint (IRIS_OTEL_ENDPOINT), which exports traces to the collector you name; the LLM judge with your own key, which sends the text it judges to that provider, and whose citation check fetches the pages an output cites; or a webhook, which posts ids, the verdict and rule names, never the text, to the address you set. No compliance certification is claimed before it is heldSOC 2, ISO 27001, HIPAA, SCIMLangfuse's page · read 2026-10-01
Cost to runFree — MIT, one process on your machine; the only spend is a judge call on a key you supply, when you opt inCloud: Hobby free (50k units a month, 30 days data access), Core $29/month, Pro $199/month, Enterprise $2,499/month; the MIT core is free to self-hostLangfuse's page · read 2026-10-01

Decision guide

Which one fits your stack?

When to choose Iris

  • You are building with MCP-compatible agents and want the integration to be one config block
  • You want the evaluation to be deterministic and local — no model call, nothing leaving the machine
  • You want to read what each rule is worth before you trust it: every rule's precision and recall is published
  • You want self-hosting to be one process and one file
  • You want a fully permissive MIT license on the whole package

When to choose Langfuse

  • You need prompt management and versioning source
  • You need LLM-as-Judge evaluation (semantic, not just heuristic) source
  • You need enterprise compliance today (SOC 2, HIPAA) source
  • You're using non-MCP frameworks and need broad SDK support source
  • You need multi-dimension dashboard slicing (user, session, model) source

FAQ

The questions buyers ask.

What is the difference between Iris and Langfuse?
Iris is an MCP-native agent eval server that needs no SDK. Add the config block, restart your client, and every session lists Iris's tools on connect. Iris never intercepts: it runs when your agent calls one of its tools, when a host hook or `iris-eval ingest` hands it a trace, or when you POST one to its HTTP API. Langfuse is an open-source AI engineering platform with native Python/JS SDKs and @observe decorators, prompt management, and 100+ library/framework integrations. Langfuse offers broader framework support and enterprise compliance (SOC 2 Type II, ISO 27001, HIPAA). source
Is Iris better than Langfuse for MCP agent evaluation?
Langfuse captures data through its SDKs, a framework integration, OpenTelemetry, or an LLM gateway such as LiteLLM, and offers LLM-as-a-Judge and Jev as a judge, code evaluators, human annotation, and enterprise compliance. source
How much does Langfuse cost to run?
Iris: Free — MIT, one process on your machine; the only spend is a judge call on a key you supply, when you opt in. Langfuse, from its own page read 2026-10-01: Cloud: Hobby free (50k units a month, 30 days data access), Core $29/month, Pro $199/month, Enterprise $2,499/month; the MIT core is free to self-host. source
Can I self-host Langfuse?
Iris: One process, one SQLite file; Docker image with a health check. Langfuse, from its own page read 2026-10-01: PostgreSQL + ClickHouse + Redis + S3 + 2 containers. source
Does Langfuse work with MCP agents?
Iris: Protocol-native — Iris is an MCP server with 12 tools; OTLP traces in. Langfuse, from its own page read 2026-10-01: MCP server in Cloud and self-hosted Langfuse: read and write tools for prompts, traces, scores, evaluators, datasets, experiments, dashboards. source

Sources

Every Langfuse statement on this page was read from one of these pages on the date shown. The file behind this page is website/src/lib/compare/langfuse.json.

Last verified: 2026-10-01. This comparison is based on publicly available documentation and may not reflect recent changes to Langfuse. We aim to keep this page accurate and fair.

See something outdated or incorrect? Report an inaccuracy — we review and update within 48 hours.

Ready to see what your agents are doing?

Add Iris to your MCP config. First trace in 60 seconds. No SDK, no signup, no infrastructure.

Install IrisTry the playgroundEvery comparison