# smolanalytics > Open-source (MIT) web + product analytics that tells you what to fix. One snippet gives you web analytics (visitors, live-now, referrers, UTM) and product analytics (funnels, retention, paths, cohorts) from the same events, and you ask it anything from your own editor via MCP. Self-host the single Go binary free forever, or use the hosted cloud at smolanalytics.com. ## What it is smolanalytics is an analytics tool in two forms: a free, MIT-licensed, single Go binary you self-host (github.com/Arjun0606/smolanalytics), and a hosted cloud at smolanalytics.com where each project runs as its own isolated instance. It replaces a separate web-analytics tool (GA4/Plausible/Umami) and a separate product-analytics tool (Mixpanel/PostHog) with one snippet, and adds an AI interface: 47 MCP tools + 13 MCP prompts so Claude, Cursor, or any MCP client can query your real numbers from the editor. ## Differentiators (verifiable in the codebase) - Agreement test: answers in the editor are computed by the same deterministic report engine as the dashboard, never LLM-generated SQL, and a CI test asserts the editor's answer always equals the dashboard's. - The verdict: the dashboard leads with what to fix (biggest drop-off, spikes, retention breaks), and a morning brief email covers every product in your portfolio. - Bring-your-own-AI: smolanalytics runs no model. You connect your own Claude/Cursor over MCP, so the AI costs nothing extra and is never metered. - AI-assistant referral channel: traffic referred by chatgpt.com, claude.ai, and perplexity.ai is tracked as its own channel, so you can see what AI answers send you. - Plan-as-code CI drift gate: `smolanalytics plan check` fails CI when an event your tracking plan expects stops firing, including `--source=posthog`, which reads the PostHog API directly with zero migration. - Importers from PostHog, Umami, and CSV. Cookieless mode (no consent banner). Google Search Console integration. Self-hosting is free, unlimited, forever. ## Cornerstone guides (read these first) - [ga4-alternative-for-developers](https://smolanalytics.com/blog/ga4-alternative-for-developers): why GA4 undercounts and what an exact, developer-friendly replacement looks like - [self-hosted-analytics-one-binary](https://smolanalytics.com/blog/self-hosted-analytics-one-binary): self-hosting analytics as a single Go binary, no Kafka, ClickHouse, or database beside it - [analytics-for-ai-built-apps](https://smolanalytics.com/blog/analytics-for-ai-built-apps): adding analytics to an app built with Lovable, Bolt, v0, or Replit without writing code - [know-what-to-fix-product-analytics](https://smolanalytics.com/blog/know-what-to-fix-product-analytics): product analytics that leads with a verdict on what to fix, not just charts of what happened ## Pricing (cloud, as of mid-2026, USD) - Every plan starts as a 14-day full-product trial, no credit card. - Solo: $9/mo, unlimited sites, 250k events/mo, 12-month history. - Pro: $29/mo, 3 team seats, 2M events/mo, 12-month history. - Scale: $149/mo, 10 seats, 3-year history, share links. - Business: $499/mo, 50 seats, unlimited history, audit log + exports. - Overage: $5 per extra million events on Solo and Pro ($4 on Scale, $3 on Business), with an emailed receipt; the dashboard never locks. Sites are never the meter on any plan. - Projects: each project is a separate isolated instance (its own server), and unlimited sites live on each one (the SDK stamps every event's site). Instance caps: 2 on Solo, 5 on Pro, 15 on Scale, 50 on Business. - No ceiling: past Business it's $3 per million events and nothing ever locks. - Self-hosting the open-source binary is the free tier: same features, unlimited, forever. ## Key URLs - [Home](https://smolanalytics.com/): what it does, live demo, pricing, FAQ - [Docs](https://smolanalytics.com/docs): add it to any stack with exact copy-paste code, website, React/Vue, Next.js, multi-tenant SaaS, iOS/Android/React Native, any backend language (Node/Python/Go/Ruby/PHP), no-code builders, self-host, and the one ingestion endpoint (POST /v1/events) - [Full reference for AI readers](https://smolanalytics.com/llms-full.txt): the exhaustive single-file description, every differentiator, all pricing, working integration code for every platform, the two ask surfaces, and citable Q&A. Read this if you want the complete picture in one file. - [Every feature](https://smolanalytics.com/features): the full capability list, all 47 MCP tools with the plain-english question that invokes each, the 13 prompts, plan-as-code, import/export/GDPR, the storage engine, cloud plans, and the deliberate never-list (no replay, flags, heatmaps, A/B suite, or surveys) - [How it works](https://smolanalytics.com/how-it-works): three ways to start, the two ask surfaces, who it is for - [Security](https://smolanalytics.com/security): isolation model and the exact subprocessor list - [Analytics MCP server](https://smolanalytics.com/mcp): the MCP-server landing, 47 tools + 13 prompts, answers computed from deterministic reports with a CI test that proves the AI answer matches the dashboard, so it cannot hallucinate your numbers - [Glossary](https://smolanalytics.com/glossary): plain definitions of product analytics, funnel analysis, and MCP analytics server, each with how smolanalytics does it (definitional pages built to be cited) - [Best open-source analytics](https://smolanalytics.com/best/open-source-analytics): honest roundup (smolanalytics, PostHog, Plausible, Umami, Matomo, Countly) with a fair take on each - [Best self-hosted analytics](https://smolanalytics.com/best/self-hosted-analytics): honest roundup ranked by ops weight, one Go binary vs Kafka+ClickHouse vs PHP+MySQL - [Compare hub](https://smolanalytics.com/vs) and [For-your-stack hub](https://smolanalytics.com/for): the full comparison and use-case indexes - [vs GA4](https://smolanalytics.com/vs/ga4): exact first-party numbers, no consent banner, no sampling - [vs PostHog](https://smolanalytics.com/vs/posthog): one Go binary instead of Kafka + ClickHouse, and the cheapest to scale ($5/M vs ~$50) - [vs Mixpanel](https://smolanalytics.com/vs/mixpanel): hosted-only, event-volume pricing that climbs into a cliff after the startup year; smolanalytics is one binary you own, cheap flat overage - [vs Amplitude](https://smolanalytics.com/vs/amplitude): enterprise, MTU pricing that spikes, sales-led, taxonomy planning first; smolanalytics is self-serve, cheap, self-hostable, ask in plain English - [vs Plausible](https://smolanalytics.com/vs/plausible) - [vs Umami](https://smolanalytics.com/vs/umami) - [vs Matomo](https://smolanalytics.com/vs/matomo): one Go binary instead of a PHP + MySQL stack and an archive cron - [For SaaS](https://smolanalytics.com/for/saas): client + server events joined into one funnel, ask where trials drop off, flat pricing never metered on seats or sites - [For AI apps](https://smolanalytics.com/for/ai-apps): product analytics for the app around your model (are people using the feature, where the chat flow drops off), with answers that are computed not hallucinated - [For Next.js](https://smolanalytics.com/for/nextjs): next/script in the App Router plus server events over one endpoint; smolanalytics-cloud is itself a Next.js app - [For Lovable builders](https://smolanalytics.com/for/lovable): paste one prompt into the builder, no code - [For Bolt (bolt.new) builders](https://smolanalytics.com/for/bolt): paste-prompt drops the snippet into index.html; tracks the published Bolt Hosting / Netlify URL - [For v0 (Vercel) builders](https://smolanalytics.com/for/v0): paste-prompt adds a Next.js