Works inside Claude, ChatGPT & any MCP agent

Statistical analysis you can actually defend.

Ask a question about your data in plain English, right in your AI chat. MCP Analytics runs real statistical computation — not a language model’s guess — and returns a validated, citable report you own and can refresh. Delivered in minutes, not the days an analyst takes.

500 welcome credits on signup — your first Brief is on us. No card required.
Price shown before you run Failed builds never billed Credits never expire
Brief · delivered in 7 min✓ Validated
Does discount depth drive repeat purchase?
orders_q2.csv · 48,212 rows · logistic regression
PredictorOdds ratio95% CIp
discount_depth1.42
<.001
prior_orders1.18
.003
email_optin1.05
.214
VIF < 2.0 Hosmer–Lemeshow p = .41 n adequate
Computed in R · glm(binomial) · citation included Refresh on new data
See how we compare to plain AI →
What you get

A piece of work, not just a report.

Calculated, not hallucinated — R computes the numbers on your data before a word of insight is written.
Ad Spend ROAS Efficiency Deck · interactive
Ad Spend ROAS report — interactive charts, sidebar navigation, AI insights
In every deliverable

Everything a defensible answer needs.

How the work gets built →
How it works

Built to order, or off the shelf.

The pipeline tailors a custom analysis to your question. The library covers the scenarios people run every day.

The pipeline · custom

Tailored to your question

The pipeline is how a custom analysis gets made — method picked, real R written and run, the result independently checked. As deep as you choose: Snapshot, Brief or Deck.

  • Custom method, code and verification
  • Three depths, priced before you run
  • Yours afterwards — re-run, modify
Understand the pipeline
or
The library · prebuilt

Proven on the common scenarios

High-quality analyses for the questions people run most — group differences, trends, churn, experiments. Verified builds, ready for your data.

  • Prebuilt and independently verified
  • A worked example with every analysis
  • Full report in minutes
Browse the library
Pricing

What’s the right answer worth?

If you want to know it’s true, that’s worth something.

How much depends on the depth of the build and the number of insights — all searchable, all maintainable.

Every tier below is an agentic pipeline — specialist agents that write and run real statistics on your data, then check the result. Not a chatbot with a nicer answer.

Asking a chatbot · Language model $0
  • ×~33% accurate on inferential statistics*
  • ×Nothing to check — no code, no method
  • ×Ask tomorrow, get a different number
Unverified
Agentic pipelines
Snapshot · Pipeline
$2
Shallow build · one insight
Your question, one chart, one verified insight computed from your own data.
  • Specialist-matched to your question
  • The R code, in the report
  • ~2 minutes
Run a Snapshot → Verified · specialised
Brief · Pipeline
$5
Deeper build · one insight, evidenced
The same answer, presented — chart, key figures, supporting evidence and the method.
  • Supporting tables where the data warrants
  • Citable — APA, MLA, Chicago, BibTeX
  • ~7 minutes
Get a Brief → Verified · specialised
Deck · Pipeline
$25
Full build · many insights
A commissioned report — multiple sections, real tests, models and diagnostics.
  • Independently verified before delivery
  • Built to your brief
  • Compared with ~$600 of consultant time
  • 30–45 minutes
Commission a Deck → Verified · specialised
Custom · Pipeline
Talk to us
Your depth · your volume
Volume, your team, priority, and deployment under your own roof.
  • Everything above, unmetered
  • Teams and shared libraries
Talk to us → Verified · specialised

EVERY TIER: COMPUTED IN R · SPECIALIST-MATCHED · THE CODE IN THE REPORT · CITABLE · YOURS TO RE-RUN AT HALF PRICE
FAILED BUILDS ARE NEVER BILLED · PRICE SHOWN BEFORE YOU RUN · *ACCURACY FIGURE: PEER-REVIEWED EVALUATION OF CHATGPT’S STATISTICAL ANALYSIS, JMIR 2025

What you get

Trusted Answers

Three things you don’t get from asking a chatbot — and the reason the answer is still useful next quarter.

01 · Understanding

Not just words, logic.

The number is computed by a statistical test on your data — the method named, the code printed in the report. The words are written after the maths, never instead of it.

How the work gets built →
02 · Repeatability

Ask the same question next quarter.

Point it at new data and it runs the identical analysis for half the price — so this quarter and last are genuinely comparable, not two different opinions.

Why the answer holds →
03 · Searchability

Every idea, indexed.

We index what each report found, not what it was called — so you can search across all of them at once, pull the cross-section that matters, and hand that same knowledge to any AI agent you use.

How semantic search works →
The alternatives

The analyst takes days. The chatbot guesses.
We compute — in minutes.

Fluent and correct are not the same thing — and rigorous doesn’t have to mean slow. Real statistical analysis used to be a choice between waiting and guessing. Now it isn’t.

Hiring it out

Days – weeks
  • ~$600 of analyst time per study
  • Back-and-forth to scope the question
  • Static PDF — new data means a new engagement

A chatbot alone

Instant — but unverifiable
  • Numbers generated by a language model
  • ~33% accuracy on inferential statistics with everyday prompts*
  • A transcript — ask again tomorrow, get a different answer

MCP Analytics

Minutes — computed & checked
  • Real R computation on your actual data
  • The right test, justified — assumptions checked, shown
  • A durable analysis you own — refresh it on next month’s data

Microsoft tells Copilot users not to rely on it for work “requiring high accuracy or reproducibility.” We exist for exactly that work.

*accuracy figure: peer-reviewed evaluation of ChatGPT’s statistical analysis, JMIR 2025
And every analysis is guaranteed. Flag anything wrong on any card — we diagnose it, fix it, and resend at no cost. Failed builds are never billed; refund requests are reviewed case by case. How refunds work →
Where it runs

Inside the AI you already use.

MCP Analytics is an MCP server — install it once and your agent can commission, run, and refresh analyses directly from conversation. No new tab, no new tool to learn.

One developer installs it. The whole team views reports in the web app. Two interfaces, one platform.

Claude Desktop ChatGPT Cursor Claude Code Any MCP Client
See the full integration guide → | API docs
Install via npx
# Add to your MCP client config { "mcpServers": { "mcpanalytics": { "command": "npx", "args": [ "-y", "@mcp-analytics/mcp-analytics", "--api-key", "YOUR_API_KEY" ] } } }

Also supports direct HTTP and OAuth2

Built for how you work: Business → Students → Analysts → Researchers → Teams & enterprise →
Try it now

Run one free, right now.

No account, no card. Upload a CSV and get a real computed report in the browser — the library’s free shelf.

Data Profile — Understand Any Dataset
The “what’s in my data” report — distributions, missingness, and structure at a glance.
Group Comparison — t-test & ANOVA
Does a numeric outcome really differ between groups? The right test, chosen and checked.
A/B Test Analysis
Upload row-level experiment data — get lift, significance, and a decision you can defend.
Time Series Forecast
Forecast any metric from its own history — with honest uncertainty bands.
Correlation Analysis
Pairwise correlations across your numeric columns — what moves together, and how strongly.
Browse all free tools →

Bring the question you’ve been sitting on.

The one you’d hire an analyst for, if it were easier. Ask it in plain English — a defensible answer comes back in minutes, and the analysis is yours for good.

Get Your First Report Free See How It Works

500 free credits on signup — your first Brief is on us · no card required