Langfuse traces your model. We measure your product and your agent.
Different jobs, and we will say so plainly: Langfuse is deeper than us on LLM tracing, evals and prompt management, and it has the bigger community. smolanalytics is one MIT Go binary that does product analytics and computed agent observability from the same events, with no extra infrastructure and no metered AI.
smolanalytics vs Langfuse, which one do I want?
It depends on what you are debugging, and plenty of teams run both. Langfuse is an MIT-licensed open-source LLM engineering platform: nested traces of every model and agent call, LLM-as-a-judge evals, datasets, and prompt management with a playground. It is deeper than smolanalytics at all three, it integrates with OpenTelemetry, LangChain, the OpenAI SDK and LiteLLM, and ClickHouse acquired it in January 2026 while committing to keep the core MIT and self-hostable. If you need trace-level LLM debugging, evals or prompt versioning, use Langfuse. smolanalytics is a different shape: a single MIT Go binary that does web and product analytics (funnels, retention, paths, cohorts, breakdowns, heatmaps, surveys, feature flags and A/B) and computed agent observability off the same event stream, tool-call volume, error rate, latency p50/p90/p99, client split, an error taxonomy, and conversation health like turns, re-ask rate, abandonment and resolution. Those numbers are computed from your events, never labelled by a model, and a CI test pins /v1, MCP and the dashboard to the same answer. Self-hosting Langfuse means Postgres, ClickHouse, Redis or Valkey and S3-compatible storage plus a web and a worker container; self-hosting smolanalytics is one binary and one file. You ask it questions in plain English from your editor: one MCP connection with an org token operates every project, 79 tools, answered by your own coding agent on your own model, so the AI is never metered by us. Pricing is a 14-day full trial with no card, Pro $49/mo for 1M events, Scale $149/mo for 10M, $8 per extra million, self-host free forever.
what each one actually is
“the leading open-source platform for LLM observability, evaluations, and prompt management”
clickhouse.com, announcing its acquisition of Langfuse, 16 Jan 2026
Self-hosting Langfuse requires Postgres, ClickHouse, Redis or Valkey and S3-compatible blob storage, plus a web container and a worker container.
langfuse.com/self-hosting, required components
smolanalytics is one MIT Go binary and one data file. Product analytics and agent observability come out of the same events: no queue, no columnar cluster, no worker to run.
what we do instead
Langfuse vs smolanalytics
| Langfuse | smolanalytics | |
|---|---|---|
| The job it is built for | LLM traces, evals, prompts | product analytics + agent health |
| Trace-level LLM debugging | deep: nested traces, playground | no, we compute rollups |
| Evals, LLM-as-a-judge, datasets | core product | not offered |
| Prompt management | versioning, deploys, playground | not offered |
| Funnels, retention, paths, cohorts | not its category | yes, same binary, same events |
| Agent numbers | traces plus model-scored evals | computed: p50/p90/p99, error taxonomy, re-ask, resolution |
| Self-host footprint | PG + ClickHouse + Redis + S3 + worker | one static binary, one file |
| Ask from your editor | MCP, project-scoped API keys | MCP, one org token routes every project, 79 tools |
| Dashboard vs API drift | separate read paths | CI pins /v1 == MCP == dashboard |
| License | MIT core, some add-ons need a key | MIT, all of it |
when to use Langfuse (often alongside us)
- You need to debug a single run call by call. Langfuse gives you the nested trace, the inputs and outputs, and a playground to replay it. We compute rollups over your events and do not store per-call LLM traces, so if the question is “what exactly happened inside that one run”, that is Langfuse, not us.
- You run evals or manage prompts. LLM-as-a-judge scoring, datasets, experiments and prompt versioning are core Langfuse products. We ship none of them and are not trying to.
- Your stack is OpenTelemetry-first, or built on LangChain, the OpenAI SDK or LiteLLM. Langfuse plugs straight into those and has the larger community and integration surface. Running both is a normal setup: Langfuse for the model layer, smolanalytics for the product and the computed agent health beside it.
Add the other half in 30 seconds
# no Postgres, no ClickHouse, no Redis, no S3, no worker docker run -p 8080:8080 -v $PWD/data:/data \ -e SMOLANALYTICS_DB=/data/smolanalytics.data \ ghcr.io/arjun0606/smolanalytics # dashboard + MCP at http://localhost:8080 # then, in your editor: "tool error rate and p90 for the last 7 days, by client"
Keep Langfuse for traces, evals and prompts. Put this next to it for funnels, retention and the computed agent numbers, on your own box, answered by your own model.
questions
Is smolanalytics a Langfuse alternative?
Only partly, and we would rather be straight about it. Langfuse is an LLM engineering platform: nested traces, evals, datasets and prompt management, and it is deeper than us at every one of those. smolanalytics is one MIT Go binary that gives you web and product analytics (funnels, retention, paths, cohorts, heatmaps, surveys, flags and A/B) plus computed agent observability from the same events: tool-call volume, error rate, latency p50/p90/p99, client split, an error taxonomy, and conversation health like turns, re-ask rate, abandonment and resolution. If the tracing and evals are what you need, use Langfuse. If you want product usage and agent health in one place with no extra infrastructure, use us. Many teams run both.
Can smolanalytics replace Langfuse for LLM tracing and evals?
No. We do not store per-call LLM traces, we do not ship LLM-as-a-judge evals, datasets or a prompt playground, and we are not planning to. Our agent observability is computed rollups over the events your app already sends, which answers “is the agent healthy and where is it failing”, not “show me every token of run 4f2a”. For that second question, Langfuse.
How are smolanalytics agent numbers different from model-scored evals?
Ours are computed, not labelled. Tool-call counts, error rate, latency percentiles, client split, the error taxonomy, turns, re-ask rate, abandonment and resolution are all derived deterministically from your event stream, so nothing is a model's opinion and nothing changes when a model changes. A CI test asserts the /v1 API, the MCP answer and the dashboard return the same number for the same question, so the answer you get in your editor cannot drift from the chart.
Is Langfuse still open source after the ClickHouse acquisition?
Yes, per their own announcements. ClickHouse announced the acquisition on 16 January 2026 and stated that Langfuse remains open source under its existing MIT license for core features, with self-hosting at production scale and the same roadmap. Some add-on features require a license key, which was already the case before. We have no reason to tell you otherwise, and this page is not an argument that Langfuse is going away.
What does smolanalytics cost, and can I self-host it?
14-day full trial, no card. Pro is $49/mo for 1M events, Scale is $149/mo for 10M, and extra usage is $8 per million. Self-hosting is free forever under MIT: one Go binary, one data file, no external services. There is no free cloud tier.
Get the verdict, not another dashboard.
Drop one snippet, ask in your editor, own your data. 14-day full trial, no credit card.