Skip to content
Go back

ProjectionLab vs cFIREsim vs FI Calc

By KingPin 13 min read
ProjectionLab vs cFIREsim vs FI Calc

Pick One Tool Per Question, Not One Tool Per Life

Use FI Calc to stress-test a withdrawal rule. Use cFIREsim when you want to model messy income adjustments against historical markets. Use ProjectionLab when you want your whole financial life in one plan, with accounts, taxes, and a timeline. That is the verdict, and the rest of this article explains why.

The tools overlap less than the “FIRE calculator” label suggests. Two of them are free simulators that replay history. One is a planning app with a free tier and a paid tier. Comparing them on price alone is like comparing a unit test to a staging environment. They answer different questions, and you will probably end up using two.

The Setup: A Scenario We Can Check

Every number below comes from this one setup, so you can reproduce it:

I did not run these three tools against each other and I won’t invent a success rate for you. The simulators run in your browser in seconds, so run this yourself. What I can do is tell you what each tool will ask for and how to read what comes back.

The 4% rule comes from Bengen’s study, which FI Calc’s docs describe as checking whether any money was left at the end, with retirements starting in 1926. The 4% figure is the rate that held up for 30-year retirements in that study. Your 40-year horizon is longer than that, which is the first reason to simulate rather than quote a rule of thumb.

Historical Simulation vs Monte Carlo (and Why You Care)

A historical simulator takes real market returns, starts a retirement in 1871, then 1872, then 1873, and so on. Each start year is a “cycle.” Your portfolio lives through the actual sequence: the Depression, the 1970s inflation, 2000, 2008. cFIREsim’s FAQ example shows 119 thirty-year cycles. Stretch the horizon to 40 years and a simulator that doesn’t loop the data loses ten start years, and the cycles that remain overlap heavily. They are not independent experiments. They are a few big events viewed from different angles, which is why sequence risk dominates the results.

Monte Carlo draws returns from a statistical model, usually using an assumed average return and volatility, and runs thousands of random sequences. You get many more paths, but they come from assumptions you typed in. Pick a 7% mean return and you get answers shaped by 7%.

Neither is correct. History shows you what actually happened, including the ugly correlations (inflation plus falling stocks) that random draws can smooth away. Monte Carlo lets you ask about futures that never occurred, such as a lower return world. If you only run one, you inherit its blind spot.

ProjectionLab’s free Basic tier lists both “Monte Carlo Simulations” and “Historical Backtesting” on its pricing page. cFIREsim and FI Calc are historical simulators. cFIREsim’s input form lists “Historical Data - All,” “Historical Data - Specific Years,” “Constant Market Growth,” and “Single Simulation Cycle” as data methods, so it also has a fixed-growth mode, but that is not Monte Carlo.

ProjectionLab: The Whole-Life Planner

ProjectionLab is the one you open when your plan has more than a portfolio in it. Kyle Nolan started building it as a solo developer in 2021. It models account types, contribution order, and drawdown, and (on Premium) cash flow and taxes. Since v4.6.0 in April 2026, its Social Security input can estimate your benefit from your Primary Insurance Amount, with a claiming age from 62 to 70 and spousal benefits.

Your plan lives in a ProjectionLab account, hosted on Google Cloud and Firebase, and you can export it to JSON. There is no Plaid and no direct bank link. You type your balances in, or a community plugin syncs them from YNAB or Lunch Money, or imports them from Monarch Money.

What it costs

As of November 2026, ProjectionLab’s pricing page lists three tiers:

ProjectionLab’s billing help page says Premium is annual only, and Pro comes monthly or annually. The pricing page shows no lifetime option. Both paid tiers offer a free trial. Checkout sits behind an account, so if the number there differs from the pricing page, trust checkout.

What the free tier gives up

The Basic tier covers forecasting, financial planning, Monte Carlo simulations, historical backtesting, and flexible modeling. Premium adds cash-flow projections, tax estimation and optimization, withdrawal strategies, Roth conversions, ACA subsidies, capital gains harvesting, compare mode, and multiple plans. The practical read: the free tier has no tax engine at all. If taxes are the question, you are in the $129 tier.

Is $129 a lot? Over 30 years it adds up to $3,870 if the price never moves, which it will. Compare that to a fund charging 0.25% more than a cheaper alternative on $1,500,000: 0.25% of $1,500,000 is $3,750 per year. A planner that talks you out of one bad decision pays for itself. A planner you open once a year does not.

Where it falls short

It is a modeling tool, not a stress-testing lab. The strength is breadth, so you spend your time building the plan, entering accounts, and setting up events. That is real work, and you need real inputs. If you want a ten-second answer to “does 4% survive 40 years,” it is slower than a free simulator.

cFIREsim: The Old Workhorse

cFIREsim.com reports version 3.0 and “Most Recent Data: Aug. 1, 2026” on its front page, so the data is current. Its front page counts more than 37 million total simulations run, which tells you people still use it.

Its withdrawal plans, as listed on the site: Inflation Adjusted, Not Inflation Adjusted, Percent of Portfolio, Variable Spending, VPW, Hebeler Autopilot, Guyton-Klinger, and Variable CAPE.

The input form is where cFIREsim earns its keep. It asks for retirement year and end year, portfolio value, initial yearly spending, inflation type (historical CPI or a flat rate), and an allocation with equities, bonds, gold, cash, and fees. It supports a glide path, a spending floor and ceiling, and up to ten adjustments, which the form describes as Social Security, pensions, or other income and expenses. It also has two Social Security entries, each with a start year, end year, and annual or monthly amount.

Its FAQ says historical stock, bond, and gold data runs from 1871 to the present, and it is blunt about taxes: “cFIREsim does not take into account taxes in any way.” That matters, because $60,000 spent is not $60,000 withdrawn if you pull from a traditional 401(k). The cFIREsim front page now also promotes FIREproof, a separate tool from the same creator that models taxes and account types.

Is it open source?

Only partly, and this is where older reviews go stale. The open-source repository (boknows/cFIREsim-open on GitHub) has a README that says the live site “has been redone, and is currently closed-source. This project will remain open as a reference.” So the code you can read is the older version, not the current live site. Do not tell anyone the current cFIREsim is open source.

Where it falls short

The interface is dense, with dozens of fields on one page. Use it when you want fine control. Skip it when you want to explain the result to a spouse.

FI Calc: The Strategy Comparison Tool

FI Calc describes itself as “a calculator that utilizes backtesting to evaluate different withdrawal strategies.” It is free, supported by donations, and says plainly that it “does not provide financial advice.”

Its configuration offers 13 withdrawal options, including Constant Dollar, Percent of Portfolio, 1/N, VPW, Guyton-Klinger, 95% Rule, CAPE-based, and Vanguard Dynamic Spending. Inputs are a starting age and duration, a starting portfolio value, an asset mix with a rebalancing frequency, an inflation-adjusted annual spending goal, and an income entry that its page says covers wages, Social Security, or other income in retirement. For historical data, the default is all available years, and simulations loop back to the first year when they run past the last one, so long retirements keep every start year. You can turn looping off. Results have Export and Share buttons.

Its historical data docs name Robert Shiller’s annual dataset, covering 1871 to 2024. Its privacy page says it processes calculator inputs in your browser, and an account is optional. An optional account saves calculations across devices, and a shared link carries your inputs in the URL, so don’t post one with your real net worth in it. The same page says Google Analytics receives page URLs, which can include those inputs.

Where it falls short

It has no account types, no tax engine, and no timeline of life events. It simulates one portfolio under one spending rule at a time. That is the point. It is the fastest way to see how a rule behaves.

Side by Side

ProjectionLabcFIREsimFI Calc
CostBasic $0/yr, Premium $129/yr, Pro $549/yr (Nov 2026)FreeFree (donation supported)
Historical backtestingYes, all tiersYesYes
Monte CarloYes, all tiersNo (fixed-growth mode)Not listed
Withdrawal strategiesPremium tier adds “Withdrawal Strategies”8 listed plans13 listed options
TaxesPremium onlyNone (per its FAQ)Not offered
Account typesYesNoNo
Social Security / incomeBenefit estimate, claiming age 62 to 70, spousalTwo SS entries plus up to ten adjustmentsIncome entry
Historical dataNot stated1871 to presentShiller, 1871 to 2024
Your dataCloud account, JSON export, no PlaidNot statedIn browser, optional account
Source codeClosedOlder reference version onlyNone found

Which Input Moves the Answer Most

Run any of these three and spending will dominate, because the withdrawal rate is spending divided by portfolio.

  1. Spending. It sets the withdrawal rate. $60,000 on $1,500,000 is 4.0%. Drop spending to $52,500 and you are at 3.5%. A $7,500 cut changes the starting rate by half a point, and no allocation tweak does that much.
  2. Horizon. Going from 30 to 40 years makes the same 4% weaker. If you retire at 45 instead of 65, you are in the 40-year (or longer) camp.
  3. Fees. A 0.50% expense ratio on $1,500,000 costs $7,500 per year. That is the entire spending cut from the previous item. cFIREsim has a fees field in its allocation, so test it.
  4. Allocation. It matters, but less than people argue about on forums. Compare 60/40 and 80/20 in FI Calc and look at which start years fail in each.

This ordering beats most hobby time. People spend a weekend choosing between two similar funds and thirty seconds on spending.

Strategy Is Where FI Calc and cFIREsim Shine

A fixed 4% rule never adjusts. Real retirees do. Guyton-Klinger cuts spending after bad years and raises it after good ones. VPW sets spending as a percentage of the remaining portfolio that rises with age. Both appear in FI Calc and cFIREsim.

The useful output is not one success number. Look at the worst starting years, the spending in those years, and how low the portfolio gets. A plan that survives only by cutting your spending 30% in year 8 is a different plan from one that never flinches. Read the spending line, not just the green percentage.

What None of Them Can Tell You

  1. Your actual future. History is one sample. A tool that says your plan worked in every cycle is describing 1871 to now, not 2027 on.
  2. Whether you can stand the dips. Seeing a 40% portfolio drop on a chart is easy. Watching it on a Tuesday while you still have a mortgage is not.
  3. Your health care bill. ProjectionLab Premium models ACA subsidies, but none of the three knows what your premiums will be. Put an explicit number in your spending, and make it bigger than you want to. Health insurance before 65 has the numbers.
  4. Your behavior. The simulator never panic sells. You might.
  5. Whether your tax number is right. Without a tax engine, a withdrawal from a pre-tax account is not a spendable dollar. Even a tool with tax estimation is only as good as the inputs you gave it.

So Which One

Start with FI Calc. It takes minutes, it is free, and it will tell you whether your spending rate is in the danger zone for your horizon. If you want to model Social Security timing, pensions, and spending floors in more detail, move to cFIREsim. If the free simulators say you are close and you need to know how your 401(k), Roth, and taxable accounts interact over 40 years, try ProjectionLab Basic first, then pay the $129 for Premium only if the tax and withdrawal-strategy features answer a question you actually have.

Do not pay for planning software to feel productive. Your 2 AM self, the one doing FIRE calculations on a Tuesday, will thank you for running three tools and noticing where they disagree. Disagreement is the useful part.

Common Questions

Is ProjectionLab worth paying for?

Yes, if you need tax modeling, Roth conversions, and withdrawal strategies. As of November 2026, ProjectionLab Premium costs $129 per year. If you only need a quick check on a withdrawal rate, the free Basic tier or a free simulator covers it.

Does cFIREsim use Monte Carlo?

No. cFIREsim replays historical market data by start year. Its data methods include all historical data, specific years, constant market growth, and a single simulation cycle. The constant growth option applies a fixed rate you enter, so it is not a random Monte Carlo draw.

Is cFIREsim open source?

Not the current site. The boknows/cFIREsim-open repository says the live cFIREsim site “has been redone, and is currently closed-source,” and that the repository remains as a reference. cFIREsim’s older code is readable there, but it is not the live version.

Is FI Calc free?

Yes. FI Calc says it is provided for free and supported by donations. FI Calc offers 13 withdrawal options, including Constant Dollar, VPW, and Guyton-Klinger, and results can be exported or shared. Its pages do not mention a paid tier.


Share this post on:

Send a Webmention

Written about this post on your own site? Send a webmention and it'll show up above once verified.


Previous Post
Net Worth Tracker: Spreadsheet vs App
Next Post
72(t) SEPP: The Other Early-Access Door

Discussion

Powered by Garrul . Sign in with GitHub or Google, or post anonymously.

Related Posts