About CalcFi

CalcFi is a free personal-finance calculator platform built on primary-source data from federal agencies. Three hundred-plus calculators. Thirty-four open datasets. Five permanent DOIs. Every formula traces to a citable government publication. No paid tier. No "expert reviewed" badges. Just the math, the source, and the work shown.

Mission

We want to help people make real money decisions with real math, real numbers, and real analysis — free, forever.

That sentence is the entire pitch. Three details inside it are non-negotiable:

• Real math. Not "estimated by our proprietary model." Not "computed by our AI." Standard, reproducible formulas from textbooks and government publications. If a CalcFi calculator gives you a number, you can take a pencil and reach the same number from the same input.
• Real numbers. Interest rates pulled live from the Federal Reserve's FRED repository. Inflation rates from BLS. Mortgage rates from Freddie Mac PMMS. Tax brackets from IRS revenue procedures. Wage data from BEA. Not synthetic. Not "indicative." Actual government-published values.
• Free, forever. The calculators do not have a paid tier. There is no premium upsell. There is no "freemium" gate. There never will be. The mission is not "build a business around personal finance calculators." The mission is "publish the math the agencies have already published, in a usable form, for anyone."

What CalcFi actually is

CalcFi at calcfi.app is a static-rendered, primarily client-side Next.js application that hosts more than 300 personal-finance calculators in nine categories:

Each calculator page has the same structure: input form at the top, computed result with all intermediate steps visible, formula displayed inline, and a citation block at the bottom listing the primary source(s) the formula derives from. Users can change inputs and watch results update; they can copy the formula; they can click through to the source publication.

Above is a live embed of the CalcFi Mortgage Payment Calculator. The iframe loads directly from calcfi.app and updates in real time. The full version includes amortization breakdown, PMI scheduling, and refi-savings projection. Try changing the rate or the term and watching the numbers re-compute.

How CalcFi differs from NerdWallet, Bankrate, SmartAsset, and similar

The honest comparison: NerdWallet, Bankrate, SmartAsset, and similar personal-finance content sites are advertising businesses with calculators bolted on. Their primary revenue model is lead generation — sending users to mortgage brokers, banks, insurance providers, tax-prep services, and financial advisors who pay per qualified lead. The calculators serve a dual purpose: deliver some real utility (which earns trust) and capture user intent at high-value moments (which earns lead-gen revenue when the user is directed to a partner).

This isn't a moral indictment. It's an accurate description of a sustainable business model. But it creates a structural conflict of interest: the calculator's job is partly to compute math, and partly to route users toward partners who pay for those users. The two jobs are not always aligned.

CalcFi has no lead-gen business model. There are no partner deals. There are no referrals to mortgage brokers, tax-prep services, banks, or advisors. There is no premium tier. The site is financed by its founder and operates with a near-zero marginal cost (Vercel free tier, Supabase free tier, primary-source data is itself free). The calculators are honest because they have nothing to sell.

Five concrete differences a user can notice on the page:

1. Citations on every formula. Calcfi pages link to IRS Pub numbers, FRED series IDs, BLS table numbers, Freddie Mac PMMS weeks, and Treasury daily yields. NerdWallet and Bankrate typically cite the government data as a brand reference, not a primary URL — they don't link to fred.stlouisfed.org/series/MORTGAGE30US for the 30-year mortgage rate they're using.
2. Formula visible. CalcFi displays the formula on the page. The user sees P × [r(1+r)^n] / [(1+r)^n − 1] for an amortizing payment, with each variable defined. Most competitors hide the formula behind "smart calculator" or "powered by AI" language.
3. No lead-gen routing. No "compare offers" button that redirects to a partner. No "talk to a financial advisor" CTA. No "find the best mortgage lender" affiliate link.
4. No "Expert Reviewed" badge. CalcFi does not display credentials of reviewers because reviewers don't add reproducibility — citations do. The page links the primary source directly.
5. Data is open, free, and downloadable. The 34 time series powering CalcFi calculators are published as CC BY 4.0 with permanent DOIs. Anyone can audit the data, fork the dataset, or build a competing calculator on the same numbers.

CalcFi Open Data — the dataset behind the calculators

CalcFi Open Data is the publicly-distributable component of the project. It is a curated, primary-source-cited collection of 34 financial and macroeconomic time series, mirrored verbatim from the original publishers — FRED, BLS, BEA, US Treasury, Freddie Mac PMMS, the World Bank, the Federal Reserve, the FDIC, and the EIA. The dataset contains 117,956 observations across daily, weekly, monthly, quarterly, and annual frequencies (preserved as published, no resampling). Coverage spans roughly 1947 through 2026 depending on series.

Series include:

• Mortgage rates: 30-year fixed (Freddie Mac PMMS, FRED MORTGAGE30US), 15-year fixed (MORTGAGE15US), 10-year fixed, 5/1 ARM
• Treasury yields: 2-year, 5-year, 10-year, 30-year, 3-month T-bill, real 10-year (TIPS), 2s10s spread (T10Y2Y)
• Federal Reserve: Federal Funds Rate, Discount Rate, Prime Rate
• Inflation: CPI-U (Consumer Price Index, BLS), Core CPI, PCE (Personal Consumption Expenditures, BEA)
• Labor: Average hourly earnings, U-3 unemployment rate, labor force participation rate, nonfarm payrolls
• Energy: WTI crude oil, Brent crude oil, US retail gasoline
• FX: USD/EUR, USD/GBP, USD/JPY
• Commodities: Copper, corn
• Macro: GDP per capita US, GDP per capita Eurozone, consumer credit, credit card APR, personal loan APR

Every series carries inline provenance — each CSV begins with comment lines stating # Source:, # License:, and # Retrieved: so the data's origin remains attached to the data itself, regardless of where the file ends up.

The dataset is permanent — registered at five DOI registrars:

• Figshare 10.6084/m9.figshare.32332290
• Zenodo (CERN-operated) 10.5281/zenodo.20302283
• OSF 10.17605/OSF.IO/PUMKT
• Kaggle DOI 10.34740/kaggle/dsv/16356447
• Mendeley Data 10.17632/jsnwhy6vjn.1

Distribution mesh

The dataset is mirrored across distinct surfaces for redundancy and discoverability. This isn't redundancy for its own sake — different audiences find data on different platforms (researchers on Zenodo, analysts on Kaggle, developers on PyPI, SQL users on BigQuery, the general public on Wikipedia-adjacent landings).

Static landing pages (8 mirrors)

• GitLab Pages — canonical landing
• GitHub Pages — open-source community mirror
• Cloudflare Pages — edge-cached global delivery
• Codeberg Pages — EU-jurisdiction, non-profit, GDPR-friendly
• Netlify — Jamstack with atomic deploys
• Surge.sh — hacker-friendly static host
• AWS S3 — enterprise infra, 99.999999999% durability
• Vercel (Datasette) — live SQL endpoint

Package registries (9 languages / ecosystems)

• Python: PyPI calcfidata · Anaconda · Read the Docs
• JavaScript: npm calcfidata
• Go: pkg.go.dev
• R: CRAN calcfidata (pending)
• Julia: CalcFiData.jl
• dbt: dbt-calcfi-open-data (hubcap PR open)
• conda-forge: staged-recipes PR #33436 (under review)

Data catalogs and warehouses (5)

• HuggingFace Datasets
• DoltHub — git-for-data, 117,956 rows versioned
• data.world
• MotherDuck — DuckDB cloud share
• BigQuery Public via Analytics Hub — calcfi_open_data listing

Interactive visualizations (23)

• 10 HuggingFace Spaces (Gradio) under @iizy
• 3 Streamlit apps: mortgage, yield curve, CPI vs PCE
• 3 Observable notebooks under @iizy
• 6 CodePen pens under collection BawbdB
• 1 Tableau Public dashboard: 30-Year Mortgage Rate History 1976-2026

Together, the mesh produces approximately 38 distinct root domains that backlink to calcfi.app — a deliberately diverse referring-domain profile chosen to make the dataset (and the calculators that consume it) durable to platform churn, censorship, and acquisition risk.

Privacy promise — your data stays in your browser

CalcFi calculators store user inputs in the browser's local IndexedDB (IDB) — they do not transmit your salary, your mortgage balance, your tax filing details, or any other personal financial information to any server. The phrase you'll see on the site is "Saves stay in your browser." We mean it literally.

The only data CalcFi collects server-side is:

• Anonymous usage analytics (page views, click events) via PostHog and Google Analytics, with IP truncation enabled and no personal identifiers attached
• Email addresses submitted explicitly (e.g., for the newsletter or saved calculator state mirroring) — but only on explicit user submit, never silently
• Decision Journal entries WITH email subscription mirror to Supabase because the cron needs to fire trigger emails — otherwise journal entries are IDB-only

We do not aggregate or sell financial-decision data. We do not run targeted ads. We do not retain IP addresses, browser fingerprints, or precise geo data in aggregate insights tables (k-anonymity ≥ 10 enforced via Supabase row-level security). The site's bar for privacy is explicit-submit-only, and the philosophy is: financial decisions are private; calculators that compute them must respect that.

Methodology and citation principles

Every calculator on CalcFi follows the same six-step methodology:

1. Identify the formula. Standard textbook or agency-published formula (e.g., the standard amortizing-loan payment formula, the IRS Pub 15 withholding tables, the BLS CPI ratio for inflation adjustment).
2. Identify the primary source for inputs. Where do interest rates come from? Where do tax brackets come from? Where do inflation rates come from?
3. Verify the source updates. Confirm cadence (FRED daily, PMMS weekly, IRS annual). Confirm the latest available value.
4. Implement the formula. Plain math. No proprietary "smart" adjustments. No interpolation, smoothing, or imputation unless the agency itself does it.
5. Display the work. Show the formula. Show intermediate values. Show the cited source.
6. Make the data downloadable. Underlying time series available in the open dataset under CC BY 4.0 with a permanent DOI.

Full methodology document at CalcFi Open Data — Methodology. The SSRN paper formalizing the citation-transparency thesis is under review.

Press placements and external citations

CalcFi and its founder have been quoted, cited, or contributed expert commentary in the following publications. Each placement is a live, indexable URL:

• NTD News — Social Security real-spending-power math
• TechBullion — Live data feeds and multi-entity consolidation
• Financial Independence Hub — ETF allocation strategy
• Freeduhm — Fintech SaaS regulatory compliance
• Marketer Magazine — Calculator-page conversion design
• CFO Drive — Finance team resilience playbooks
• The Business Navigator (LinkedIn) — Salary thresholds and paycheck math

Team and founder

CalcFi is built and operated by Jere Salmisto, based in Helsinki, Finland. The full founder profile, work history, expertise areas, and contact information lives at About Jere Salmisto.

The project is intentionally a solo, sustainable, low-overhead operation. There are no employees, no investors, no equity to dilute. The financial logic is straightforward: hosting costs are sub-$50/month on free tiers, the primary-source data is free, the calculators are static-rendered. CalcFi can run indefinitely without monetization because it does not require monetization to operate.

Contact and partnership

For press inquiries, data licensing, embed partnerships, integration questions, or research collaboration:

• Email via the CalcFi about page
• Embed the open dataset in your tools: see CalcFi developers page
• Cite the dataset academically: see the methodology document for preferred citation format
• Featured.com / HARO / Connectively / Qwoted journalists: typically same-day response on personal-finance, mortgage, tax, retirement, or fintech topics

CC BY 4.0 + MIT. The CalcFi Open Data dataset is licensed under Creative Commons Attribution 4.0 International. Code (Python, R, JavaScript, Go, Julia, dbt packages) is licensed MIT. Use it commercially. Use it noncommercially. Build a competitor on top of it. The license is permissive on purpose — the goal is widespread use, not artificial scarcity.