Refinance Comparison Calculator

What the calculator actually compares

The calculator amortizes your current loan (balance, rate, and remaining months as you enter them) to get a current monthly payment and a current total-interest figure. It then amortizes each of your two offers separately — each with its own rate, term, closing costs, and discount points — and compares every offer back to that current-loan baseline.

For each offer, discount points are converted to dollars the standard way: one point costs 1% of the balance being refinanced, so on a $400,000 balance one point is $4,000. That points cost is added to the closing costs you enter to get a single total closing cost figure for the offer, which is what the break-even math below is measured against.

A worked comparison at this week's rates

Suppose your current loan is a $400,000 balance at 6.50% over 30 years — the Freddie Mac Primary Mortgage Market Survey 30-year average from exactly one year earlier. That loan's own amortization schedule gives a current payment of $2,528 and total interest, if carried to term, of $510,178.

Offer A is strictly worse on both measures: the 30-year fixed rate went up over the past year, so refinancing into it raises both the monthly payment and the lifetime interest bill. Offer B is the more interesting case — it saves $301,044 in interest over the life of the loan, but only by raising the required payment by $856 a month for fifteen years. Neither of those is a mistake in the calculator; they are two different, both-correct ways of reading the same rate environment.

A real gap in the break-even figure worth knowing about

The calculator's break-even month is defined as total closing costs divided by monthly savings, rounded up. That formula only makes sense when monthly savings is positive. When it isn't — as in both offers above, since both raise the monthly payment — the calculator does not report 'no break-even' or leave the field blank. It returns 0.

The same logic explains why Offer B's headline net savings can be positive (net of the $301,044 interest figure above any closing costs entered) even though its break-even month reads 0. Net savings is a lifetime figure; break-even is a monthly-cash-flow figure. They are answering different questions and, as this example shows, they can point in opposite directions on the same offer.

Why comparing total interest across different terms needs care

The 'interest saved' figure above compares total interest on your current loan's schedule against total interest on the new loan's schedule — but those two schedules don't necessarily cover the same span of time. A 15-year offer measured against a 30-year current loan is comparing 180 months of new payments to whatever months remained on the old loan, which in the current example is the full 360. The $301,044 interest-saved figure is real, but it partly reflects paying the loan off in half the time, not the rate difference alone — the same trade-off the 15-year-versus-30-year comparison makes explicit. A dollar-for-dollar apples comparison would hold the payoff date constant across offers; this calculator does not do that automatically, so read a large interest-saved number on a shorter-term offer with that in mind.

How 'the better offer' gets decided

The calculator labels one offer as better only when their total costs — total interest plus total closing costs — differ by more than $2,000. Inside that $2,000 band, it reports the two offers as 'similar' rather than picking a winner. That threshold is a fixed constant in the tool, not something either offer's numbers can move. In the worked example, Offer A and Offer B differ in total cost by well over $300,000, so the labeling has no practical effect here — but on two closely priced offers, a marginal difference just outside the $2,000 line will get a definitive verdict where one just inside it will be called a toss-up.

Reading a real Loan Estimate against this calculator

Two things on an actual refinance offer don't show up in the three inputs this calculator takes for each offer, and both change the real cost.

• ·Cash-out proceeds are not automatically deductible. Per IRS Publication 936, no matter when the debt was incurred, you can no longer deduct interest on the portion of a loan whose proceeds were not used to buy, build, or substantially improve the home securing it. A cash-out refinance that pulls equity for anything else — debt consolidation, tuition, a car — does not generate deductible interest on that cashed-out slice, even though the whole new balance is one loan with one rate in this calculator.
• ·The interest deduction is capped independent of your rate. The same publication limits the mortgage interest deduction to the first $750,000 of acquisition debt ($375,000 married filing separately), or $1,000,000 for debt from before December 16, 2017. A large refinance balance above that line pays non-deductible interest on the excess regardless of which offer you pick.

Why the same rate quote doesn't always mean the same pricing

This calculator treats 'the rate' on each offer as a single flat input, but the rate a lender actually quotes you depends in part on whether your refinanced balance sits inside or outside that year's conforming loan limit. The 2026 baseline conforming limit set by the FHFA is $832,750 for a one-unit home, rising to a high-cost-area ceiling of $1,249,125. A refinance balance that crosses that line prices as a jumbo loan, and jumbo pricing does not automatically track the conforming-loan averages this page's worked example is anchored to. If either of your two offers refinances a balance above the applicable limit for your county, treat the 6.71%/6.04% figures above as a conforming-market reference point, not as your quote.

The same conforming-limit boundary matters for a different reason on a cash-out refinance specifically: increasing your balance through cash-out can push a loan that was comfortably conforming over the line, changing both the rate you're quoted and, per the deduction rule above, how much of the resulting interest is deductible at all.