--- title: Langfuse vs. Raindrop shortTitle: Raindrop seoTitle: "Raindrop Alternative: Langfuse vs. Raindrop" description: Langfuse is the open-source Raindrop alternative. October 2026 comparison of self-hosting, ClickHouse, unit vs event pricing, retention, and experiments. tags: [comparison] --- # Langfuse vs. Raindrop This guide outlines the key differences between **[Langfuse](/)** and **[Raindrop](https://www.raindrop.ai/)** to help engineering teams choose the right AI engineering platform. All Raindrop facts were checked against public Raindrop pages in October 2026. ## TL;DR - **Open-source core, free to self-host.** Langfuse's core tracing, evaluation, prompt management, and APIs are [MIT-licensed](#open-source-and-deployment) and available on Cloud or self-hosted; additional Enterprise features require a license. Self-hosting is free on every tier, with Enterprise SSO and organization RBAC included. Raindrop's monitoring platform is a commercial hosted product: Workshop, the MIT-licensed part, is a local debugger, and self-hosting is advertised as a beta. - **About 65% lower on the illustrated workload, with no seat fees.** 500,000 user turns at 12 Langfuse units each is $621 on Langfuse Pro where [Raindrop Pro is $1,799](#worked-example): a $299 base plus $0.003 per event. Both figures are subscription plus telemetry usage; model inference and add-ons are separate. Langfuse Pro includes 3 years of history and unlimited users; Raindrop Pro includes unlimited retention. - **Data through APIs and exports, or SQL when self-hosting.** Langfuse runs on ClickHouse, [open source since 2016](#storage); read it through the Observations and Metrics APIs, scheduled exports to your bucket on Pro with the Teams add-on, Enterprise, and self-hosted, or SQL on your own ClickHouse. Raindrop data export starts on Pro, and warehouse exports are Enterprise. - **The improvement loop in one project.** Langfuse runs [prompt and model experiments in the UI](#improvement-loop) and complete agents through the SDK, compares results to a baseline, and keeps annotation queues, prompt labels, and CI gates in the same project. Raindrop's hosted experiments compare traffic already logged and do not run or replay the agent. - **Same-day switch.** [Point the SDK or OpenTelemetry exporter at Langfuse](#switching-from-raindrop) and import dataset cases via CSV, SDK, or API. Most teams see first traces in Langfuse the same day. **Choose Langfuse** if you want to own the platform: an open-source core for production traces, datasets, experiments, annotation, and prompts, on Cloud or self-hosted including air-gapped, on ClickHouse you can query and export. One meter covers unlimited users, and on the [worked example](#worked-example) it is about 65% lower than Raindrop Pro at list prices. Proven at 21 of the Fortune 50. **Choose Raindrop** if you are already running it and the differences on this page do not justify a switch. You accept a commercial hosted platform with self-hosting in beta, a $299 base plus per-event pricing, a Triage message allowance on Pro, warehouse export on Enterprise, and hosted experiments that compare logged traffic rather than running the agent. --- ## Open source and deployment [#open-source-and-deployment] Langfuse's core tracing, evaluation, prompt management, and APIs are MIT-licensed and available on Cloud or self-hosted; additional Enterprise features require a license. [Self-hosting](/self-hosting) runs the UI, login, and data in your environment on every tier, including [air-gapped operation](/self-hosting/security/networking), and Enterprise SSO and organization-level RBAC ship in the [free distribution](/pricing-self-host). Langfuse Cloud is covered by [SOC 2 Type II and ISO 27001 audits, is GDPR compliant](/security), and signs a BAA from the Pro plan in the [HIPAA region](/security/hipaa). Langfuse is used by [21 of the Fortune 50 and 129 of the Fortune 500](/enterprise). Raindrop's monitoring platform is a hosted, paid product. [Workshop](https://github.com/raindrop-ai/workshop), the MIT-licensed part, is a local debugger, and self-hosting is advertised as a beta. Its [security docs](https://raindrop.ai/docs/security/data-security-and-privacy) place event storage in AWS with United States subprocessors. | | Langfuse | Raindrop | | -------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------- | | License | **[MIT](https://github.com/langfuse/langfuse)** core: tracing, evaluation, prompts, APIs; Enterprise features licensed separately | MIT for Workshop; the platform is commercial | | GitHub stars | [![Langfuse GitHub stars](https://img.shields.io/github/stars/langfuse/langfuse?style=for-the-badge&label=%20&labelColor=black&color=orange)](https://github.com/langfuse/langfuse) | Workshop only | | Full-stack self-hosting | **[Every tier, including free](/pricing-self-host)** | Beta | | Enterprise SSO (self-host) | **Included in free OSS** ([self-hosted pricing](/pricing-self-host)); Cloud via the [Teams add-on](/pricing) | Enterprise | | Org-level RBAC (self-host) | **Included in free OSS** | Owner, Admin, Member roles | | Air-gapped operation | Supported | Not described | | Data regions | Cloud EU, US, Japan; [HIPAA region](/security/hipaa); any region via self-hosting | AWS, United States | | SOC 2 | [Type II](/security), reports on Pro | Type II | | Published Cloud Enterprise | **[$2,499/month](/pricing)** | Custom, sales-gated | --- ## Storage and data access [#storage] [ClickHouse](https://clickhouse.com/) has been [open source since June 15, 2016](https://clickhouse.com/blog/open-source-10) (Apache 2.0) and is today the most popular open-source analytical database, with 2,000+ contributors. Langfuse Cloud and self-hosted run the same ClickHouse architecture, so you can move between them without locking into a closed store. Langfuse works directly with ClickHouse's core team as [part of ClickHouse](/blog/joining-clickhouse). Every LLM call, tool execution, and agent step is written once to a wide, immutable observations table ([details](/blog/2026-03-10-simplify-langfuse-for-scale)). High-volume reads go through the [Observations](/docs/api-and-data-platform/features/public-api#v2) and [Metrics](/docs/metrics/features/metrics-api#v2) APIs, [scheduled exports](/docs/api-and-data-platform/features/export-to-blob-storage) (Pro with the Teams add-on, Enterprise, and self-hosted) land JSONL or Parquet in your own S3, GCS, or Azure bucket, and self-hosters query ClickHouse in SQL. Raindrop's store is vendor-operated. Reads go through its Query API and MCP; data export is a Pro feature on the [plans page](https://raindrop.ai/docs/platform/plans), and the pricing cards put Snowflake and BigQuery exports on Enterprise. | | Langfuse | Raindrop | | --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------- | | Data store | **ClickHouse**, [launched 2016 (OSS)](https://clickhouse.com/blog/open-source-10), Apache 2.0 | Vendor-operated, on AWS | | Cloud and self-hosted | Same ClickHouse engine and observations-first data model ([v4](/docs/v4)) | Self-hosting in beta | | Getting data out | SQL on ClickHouse when self-hosting; Observations and Metrics APIs; [scheduled export](/docs/api-and-data-platform/features/export-to-blob-storage) to your bucket (Pro with Teams add-on, Enterprise, self-hosted) | Query API and MCP; export on Pro; Snowflake / BigQuery on Enterprise | | Retention | 30 days Hobby, 90 days Core, **3 years** Pro and Enterprise | 14 days Hobby, unlimited on Pro | --- ## Pricing [#pricing] Raindrop bills **events**. The pricing card defines one event as a user message and agent response, including tool calls and sub-agents, or one background agent run. Langfuse bills [units](/docs/administration/billable-units): one trace, one observation, or one score. A single Raindrop event can contain many Langfuse observations, so the cheaper bill depends on trace shape. The table and example below cover the software subscription and telemetry usage only. Model inference for judges or Triage, privacy add-ons, and export add-ons are separate on both sides. | | Langfuse | Raindrop | | ------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ | | Free tier | 50,000 units/month, 2 users, 30 days | 1,000 events/month, hard stop, 14 days | | Paid entry | **Core $29/month** (100,000 units, unlimited users, 90 days) | **Pro $299/month** plus usage | | Usage | $8/100k units after 100k, then $7/100k from 1M to 10M ([pricing](/pricing)) | **$0.003/event** up to 1M events, then **$0.002/event** | | Retention | 30 days Hobby, 90 days Core, **3 years** Pro and Enterprise | 14 days on Hobby, **unlimited** on Pro | | Seats | No seat fees on Core and above | None published | | In-app agent | [Halo](/docs/halo) on every Cloud plan, usage limits by plan; investigates traces and creates datasets, prompts, and dashboards | Triage: 500 messages/month on Pro, more purchasable; unlimited on Enterprise | | PII | SDK-side [masking](/docs/observability/features/masking) with rules you write, applied before data leaves your app; included | Managed detection and redaction via the [PII Guard](https://raindrop.ai/docs/security/pii-redaction) add-on, from $80/month plus per-event tiers | | Export | [Scheduled blob export](/docs/api-and-data-platform/features/export-to-blob-storage) on Pro with the $300/month Teams add-on, Enterprise, and self-hosted | Pro; Snowflake / BigQuery on Enterprise | Masking and PII Guard are different approaches, not the same feature at two prices: Langfuse masks in your SDK with rules you define before data is sent; Raindrop's PII Guard detects and redacts on its side after ingestion. ### Worked example [#worked-example] 500,000 user turns/month · 10 observations per turn · 1 score per turn. Each turn is one Raindrop event; in Langfuse it is 1 trace + 10 observations + 1 score = **12 units**. | | Base | Usage | **Total/month** | | ---------------- | ------- | --------- | --------------- | | **Langfuse Pro** | $199.00 | $422.00 | **$621.00** | | **Raindrop Pro** | $299.00 | $1,500.00 | **$1,799.00** | Public list prices. About 65% lower on Langfuse Pro in this illustrated workload: $621 versus $1,799 per month. - **Langfuse Pro:** 6,000,000 units. 100k included, 900k at $8/100k ($72), 5,000,000 at $7/100k ($350). Base $199. **3-year** data access. - **Raindrop Pro:** 500,000 events at $0.003. Base $299. **Unlimited** retention. - **Excluded on both sides:** model inference for judges or Triage, privacy add-ons (Raindrop PII Guard), and export add-ons (the Langfuse Teams add-on for scheduled bucket export). Raindrop bills the turn; Langfuse bills every observation and score inside it. A lean turn is cheaper on Langfuse Pro; a turn with dozens of model calls, tool spans, and scores can be cheaper on Raindrop Pro. See [Langfuse pricing](/pricing), [start free](https://cloud.langfuse.com), or [talk to us](/talk-to-us) about Enterprise volumes. --- ## The improvement loop [#improvement-loop] Langfuse runs the loop in one project. Add a production observation to a [dataset](/docs/evaluation/experiments/datasets), then run an experiment: [prompt and model experiments](/docs/evaluation/experiments/experiments-via-ui) execute prompts and models in the UI, and a complete agent runs through an [SDK experiment](/docs/evaluation/experiments/experiments-via-sdk) in your repository with your existing mocks and test environment, with results [compared to a baseline](/docs/evaluation/experiments/compare-experiments) in Langfuse. Route outputs to [annotation queues](/docs/evaluation/evaluation-methods/annotation-queues). [Prompt labels](/docs/prompt-management/features/prompt-version-control) promote a version without a redeploy, and the version can carry [model and tool config](/docs/prompt-management/features/config) that your application reads at runtime. [LLM-as-a-judge](/docs/evaluation/evaluation-methods/llm-as-a-judge) and [code evaluators](/docs/evaluation/evaluation-methods/code-evaluators) score live observations and experiment runs; judges use a model connection you choose. In production, [Pulse](/docs/observability/features/pulse) surfaces outliers, [alerts](/docs/observability/features/alerts) fire on thresholds you define, and [Halo](/docs/halo) investigates traces and creates datasets, prompts, and dashboards on request; it is on every Cloud plan and in public beta for self-hosted. Raindrop's docs describe the product as monitoring for silent agent failures: issue detection, custom signals, a Triage agent, datasets, and A/B tests. Its hosted [experiments](https://raindrop.ai/docs/platform/experiments) compare cohorts of logged events and, in Raindrop's words, "don't run or replay your agent"; Workshop and Simulations provide separate replay and testing workflows. | | Langfuse | Raindrop | | ------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------- | | Finding a failure in production | Judges and code evaluators on live observations, Pulse, alerts, Halo | Issue detection (Pro), signals, Triage | | From a failure to a test | [One-click add to a dataset](/docs/evaluation/experiments/datasets), then a prompt experiment in the UI or an SDK experiment of the agent | Add to dataset; hosted experiments compare logged cohorts | | Who runs the comparison | Prompts and models: the Langfuse UI. Full agents: your SDK run, compared in Langfuse | Hosted experiments: logged traffic. Replay: Workshop or Simulations | | Human review | [Annotation queues](/docs/evaluation/evaluation-methods/annotation-queues); unlimited on Pro | Labels on dataset items | | Prompt release | [Labels](/docs/prompt-management/features/prompt-version-control) with versioned config your app applies at runtime | Not documented | | CI | [GitHub Action](/docs/evaluation/experiments/experiments-ci-cd) on experiment results | Via the separate Simulations product | --- ## Which should you choose [#which-should-you-choose] Choose **Langfuse** if you want **an open-source platform that works across your stack**: OpenTelemetry instrumentation for any framework, free self-hosting with SSO and organization RBAC included, prompt experiments in the UI and full-agent experiments through the SDK against a curated dataset, annotation queues and prompt labels on the same traces, data in ClickHouse you can query and export, and unit pricing with no seat fees. Choose **Raindrop** if you are **already running it** and the differences above do not justify a switch. --- ## Why teams move away from Raindrop [#why-teams-move-away] The same reasons recur in migration conversations. - **Event pricing with a $299 floor.** Raindrop Pro starts at $299 and charges [$0.003 per event](#pricing), so the bill scales with every user turn. On the [worked example](#worked-example) that is $1,799 on Raindrop Pro versus $621 on Langfuse Pro, which meters units with no seat fees and includes 3 years of history. The gap shrinks when one event contains dozens of observations. - **The platform you buy is hosted; the open-source artifact is a local debugger.** Raindrop's Workshop is free and local, but the monitoring platform teams pay for is the hosted plan, with [self-hosting in beta](#open-source-and-deployment) and warehouse export on Enterprise. Langfuse self-hosts its open-source core, UI, login, and evaluators included, on [ClickHouse you can query](#storage). - **Hosted experiments do not run the agent.** Raindrop's hosted experiments [compare logged traffic](#improvement-loop) and do not run or replay the agent. Teams that need a candidate scored against a curated set run prompt experiments in the Langfuse UI or a full-agent experiment through the SDK, then promote a prompt label from the result with a CI gate. - **Triage is metered in messages.** Raindrop Pro includes a Triage allowance of 500 messages a month, with unlimited on Enterprise. Compare it separately with [Halo's plan-based usage limits](#pricing); evaluator execution and threshold alerts are different workloads and are not a substitute for either. --- ## Why teams choose Langfuse [#customers] A pattern repeats across production adopters: start where governance requires, often self-hosted, then grow into the platform. - **[Canva](/users/canva)** runs a multi-agent support system for a product with **250 million** monthly active users. LLM-as-a-judge evaluators score the systems across **15–20** metrics, and domain experts build evaluators without an engineering handoff. They self-hosted first, then moved to Cloud. - **[SumUp](/users/sumup)** rolled AI support to **35+** markets and deflected almost **50%** of conversations, cutting external BPO cost by **30%**. They started self-hosted for the proof of concept, then moved to Langfuse Cloud. - **[Merck](/users/merckgroup)** runs about **80** of 300+ GenAI use cases on Langfuse, with **200+** people building on it. They self-host for data sovereignty and provision projects via API. --- ## Switching from Raindrop [#switching-from-raindrop] Move live instrumentation and recreate the durable assets. History stays readable in Raindrop for the plan's retention window. Most teams see first traces in Langfuse the **same day**. - **Instrumentation:** initialize the [Langfuse SDK](/docs/observability/sdk/overview) or send OpenTelemetry to the [OTLP endpoint](/integrations/native/opentelemetry). Run both exporters during the overlap. - **Datasets, signals, and prompts:** export a Raindrop dataset as CSV and import it through the [Langfuse CSV uploader, SDK, or API](/docs/evaluation/experiments/datasets). Recreate judges and [prompts](/docs/prompt-management/get-started) in Langfuse. A Raindrop signal is a detector on logged events; the matching Langfuse object is an evaluator or a score. - **Trace history:** stays in Raindrop. Pro's unlimited retention means the archive can remain there while new data lands in Langfuse. [Talk to us](/talk-to-us) if you want help planning the overlap. --- ## Start free: Cloud or self-host [#get-started] [Langfuse Cloud](https://cloud.langfuse.com) Hobby includes 50k units/month, no credit card. Or [self-host](/self-hosting) the MIT stack. The [example project](/docs/demo) is populated before you instrument anything. - [Get started free](https://cloud.langfuse.com) - [Talk to us](/talk-to-us) --- ## FAQ [#faq] ### Is Langfuse a Raindrop alternative? [#is-langfuse-a-raindrop-alternative] Yes, if you want an open-source core for tracing, datasets, experiments, human review, and prompt management on Cloud or self-hosted, with unit pricing and no seat fees. See [Which should you choose](#which-should-you-choose). ### Is Raindrop open source? [#is-raindrop-open-source] Workshop is, under the MIT license. It is a free, local debugger. The hosted monitoring platform is a paid product, and self-hosting is advertised as a beta. Langfuse's core tracing, evaluation, prompt management, and APIs are MIT-licensed; additional Enterprise features require a license. See [Open source and deployment](#open-source-and-deployment). ### How does Raindrop pricing compare to Langfuse? [#how-does-raindrop-pricing-compare] Raindrop Pro is $299/month plus $0.003 per event up to 1 million events, then $0.002. An event includes the tool calls inside one user turn. Langfuse Pro is $199/month plus graduated unit pricing, and a score or tool span is its own unit. At 500,000 turns with 12 Langfuse units each, the list-price totals are **$621** on Langfuse Pro and **$1,799** on Raindrop Pro, subscription plus usage, with model inference and add-ons excluded. Raindrop Pro includes unlimited retention; Langfuse Pro includes 3 years. Details in the [worked example](#worked-example). ### Can Raindrop Simulations use Langfuse traces? [#can-raindrop-simulations-use-langfuse-traces] Raindrop's [Simulations page](https://www.raindrop.ai/simulate/) lists Langfuse as a bring-your-own-traces source, next to Helicone, so Langfuse stays the system of record for a team that uses both. ### Do Raindrop experiments run the agent? [#do-raindrop-experiments-run-the-agent] No. Raindrop's docs state its hosted experiments compare logged traffic and do not run or replay the agent; Workshop and Simulations are separate replay and testing workflows. Langfuse prompt experiments execute prompts and models in the UI, and a complete agent runs through an SDK experiment in your repository with results compared in Langfuse. See [The improvement loop](#improvement-loop). ### How do I switch from Raindrop to Langfuse? [#how-do-i-switch-from-raindrop] Run both in parallel, add the Langfuse SDK or repoint OpenTelemetry, and import dataset cases. Trace history stays in Raindrop. Most teams see first traces the same day. See [Switching from Raindrop](#switching-from-raindrop). [Ask another question](/docs/ask-ai) This comparison is out of date? Please [raise a pull request](https://github.com/langfuse/langfuse-docs/blob/main/content/compare/raindrop.mdx) with up-to-date information. **Start with Langfuse** [Sign up](/cloud) --- ## Agent Instructions This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/compare/raindrop.md`. ### Querying these docs If the answer is not on this page, query the documentation instead of guessing: - **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question: ```bash curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?" ``` - **Index of every page**: , with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt). ### Before writing Langfuse code - **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results. - **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application. - **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples. - **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx @langfuse/cli api `) to read or write traces, prompts, datasets, and scores from the terminal. Found an error in these docs? Please open an issue at .