SwitchWize decision guide

Capital One Savor or American Express Gold: is the annual fee worth it for dining and groceries?

This is the coupon-book question. Add up every Amex credit at face value and a fee card looks like a bargain; count only what you would have bought anyway and the arithmetic changes completely. So every credit starts at zero here and counts only once you tick it. The two currencies are also not alike: cash back is worth its face value by contract, while a point is worth whatever you can reliably redeem it for, which is your judgement rather than a fact.

SwitchWize Research DeskUpdated August 15, 2026Data checked August 15, 20268 min read

What you can expect

  • No login required
  • No credit pull
  • Assumptions shown
  • Sources included

Quick answer

Tick only the credits you would genuinely use, then read the spend boundary. If your dining and grocery spend is below it, the no-fee card is ahead and there is nothing to remember each month.

If savor

Better below the spend boundary, and better for anyone who does not want to remember a monthly credit. Cash back is worth its face value with no redemption decisions.

If amex gold

Better above the spend boundary, or when the credits you would genuinely use already cover most of the fee.

Key number to watch

Required fact is missing: savor_dining_cashback_rate

How we calculated this

Test your situation

See your result in dollars

Change any number below to match your situation. No login is required, and your entries stay in this browser.

$

Restaurants and takeaway. One of the two categories that decides this.

$

Supermarkets. Warehouse clubs and superstores often do not count.

$

Tickets and events, where the cash-back card earns and the fee card does not.

Your assumption, never a reported fact. It moves the answer more than any other input.

Your answer so far

The card terms we hold conflict or are out of date. Naming a winner from those would be worse than saying so.

See the full breakdown

The card terms we hold conflict or are out of date. Naming a winner from those would be worse than saying so.

We are not showing a winner for this one

  • Required fact is missing: savor_dining_cashback_rate
  • Required fact is missing: savor_grocery_cashback_rate
  • Required fact is missing: savor_annual_fee
  • Required fact is missing: amex_gold_dining_earn_rate
  • Required fact is missing: amex_gold_grocery_earn_rate
  • Required fact is missing: amex_gold_annual_fee

Capital One Savor

Not modeled

modeled after-tax interest

American Express Gold

Not modeled

modeled after-tax interest

Try a scenario

What could change this

Required fact is missing: savor_dining_cashback_rate

How certain: scenario dependent

What matters most

Most ongoing annual value

not applicable

Required fact is missing: savor_dining_cashback_rate

Side-by-side comparison

Annual fee

Savor
None
Amex Gold
A meaningful annual fee

What you earn

Cash back is contractual. A point valuation is a judgement.

Savor
Cash back at a fixed rate
Amex Gold
Membership Rewards points, worth what you can redeem them for

Dining and groceries

Savor
A flat cash-back rate
Amex Gold
A higher points rate, capped on annual spend

Credits

Counted here only when you say you would use them.

Savor
None to remember
Amex Gold
Monthly credits at specific merchants

Effort

Savor
None
Amex Gold
Monthly credits and redemption decisions

What could go wrong

Capital One Savor

What needs to work
Your dining and grocery spend stays below the boundary where the fee card catches up.
Common problem
Assuming the whole grocery bill earns the bonus rate when much of it is at a superstore.
What it could cost
The extra earning you gave up if your spend is well past the boundary.
How to prepare
Easily reversible, with no fee at risk.

American Express Gold

What needs to work
You spend past the boundary, or you genuinely use enough credits to cover the fee.
Common problem
Counting every credit at face value in January and using half of them by December.
What it could cost
A full annual fee against credits you did not use and points you could not redeem well.
How to prepare
Reversible, but the fee is charged before you know how the year goes.

A simple backup plan

Tick the credits first, read the boundary second

The order matters, because a face-value credit total makes any fee look worth paying.

  1. 1Start with no credits counted, which is the default here.
  2. 2Tick only the ones you would genuinely use, and enter what you would have spent there anyway rather than the face value.
  3. 3Subtract the superstore and warehouse-club share from your grocery figure, because it usually earns the base rate.
  4. 4Read the spend boundary: the dining and grocery spend at which the fee card catches up on earning alone.
  5. 5Set the point value to something you can reliably redeem at, and watch how far the boundary moves when you change it.

Educational illustration only. The right amount depends on your needs and timing.

What to do next

  1. Question 1

    Would you actually use the monthly credits?

    Yes: Enter what you would have spent there anyway, not the face value.

    No: Count them as zero, which is what this page does by default.

  2. Question 2

    Is your dining and grocery spend past the boundary?

    Yes: The fee card is ahead on earning alone.

    No: The no-fee card wins, and there is nothing to remember.

  3. Question 3

    Can you reliably redeem points above a cent?

    Yes: The boundary moves in the fee card’s favour.

    No: Use one cent and see whether the answer still holds.

Plain-text decision tree. Would you actually use the monthly credits? If yes, Enter what you would have spent there anyway, not the face value. If no, Count them as zero, which is what this page does by default. Is your dining and grocery spend past the boundary? If yes, The fee card is ahead on earning alone. If no, The no-fee card wins, and there is nothing to remember. Can you reliably redeem points above a cent? If yes, The boundary moves in the fee card’s favour. If no, Use one cent and see whether the answer still holds.

When to check again

  • Either card changes its category rates, caps or annual fee.
  • Amex changes which credits it offers, or the merchants they apply to.
  • Your dining or grocery spend changes materially.
  • You stop using a credit you had been counting on.

Methodology

Savor cash back is priced from contractually reported category rates at one cent on the dollar, unless the issuer’s own terms say otherwise. Amex points are valued at your own figure, which is recorded as an assumption and never reported as fact. Category caps apply to spend, so spend above a cap earns the base rate rather than the headline rate. Grocery spend at merchants outside the category earns the base rate on both cards. Every Amex credit is zero until you tick it, and is then worth face value times the share you use times the share that replaces something you would have bought anyway, never more than you would naturally have spent. Welcome offers are excluded from ongoing value. The break-even spend is solved rather than searched, so it responds smoothly as the point value changes.

Card terms are reported facts with an expiry. Contradictory or expired terms abstain rather than picking one.

  • Transfer-partner redemptions above a cent are possible and are not assumed. If you cannot repeat a redemption reliably, it is not a value.
  • Purchase protection, insurance and status benefits carry no dollar figure here.
  • Welcome offers are excluded entirely from the ongoing comparison.
  • Merchant category coding varies, and a shop that looks like a supermarket may not code as one.
  • This is education, not credit advice, and neither card is recommended on its rewards alone.

Card terms re-checked on every page build; the guide reviewed monthly. Editorial conclusions do not depend on affiliate availability.

Sources

Frequently asked questions

Is the Amex Gold annual fee worth it?

It depends on two things and this page separates them. First, the credits: they are worth nothing here until you tick them, and then only what you would have spent at those merchants anyway. Second, the spend boundary: the annual dining and grocery spend at which the extra earning recovers the fee on its own. Below that boundary with no credits ticked, a no-fee cash-back card is simply ahead.

How much dining and grocery spend do I need to justify the fee?

That is the number this page leads with. It is solved from your own inputs rather than quoted from a generic example, and it moves as soon as you change the point value, tick a credit, or move grocery spend to a superstore that falls outside the category.

Why are the Amex credits set to zero by default?

Because adding them up at face value is how a fee card comes to look free. A credit is worth what you would have bought anyway at that merchant in that month, and a credit you forget in March is gone rather than carried to April. Starting at zero and adding back what you would genuinely use is the honest direction to work in.

Is cash back worth the same as points?

No, and this page does not treat them alike. Cash back is worth one cent on the dollar by contract, so it is a reported fact. A point is worth whatever you can reliably redeem it for, so it is your assumption and is recorded as one. Treating a point as cash is the other common way a fee card gets overstated.

Does all my grocery spend earn the supermarket rate?

Often not. Superstores and warehouse clubs usually code outside the supermarket category and earn the base rate on both cards. There is a separate input for that spend here, because leaving it in the grocery figure overstates whichever card has the higher supermarket rate.

What happens when I spend past the category cap?

Spend above the annual cap earns the base rate rather than the headline rate. The cap applies to spend rather than to the dollars earned, which is the correct way round: capping the dollars afterwards would quietly pay the bonus rate on spend that never earned it.

Should I switch if my current card is already fine?

Not necessarily, and this page will tell you so. Enter what your current card is worth and the gain that would make an application worthwhile, and if neither card clears it the answer is to stay put rather than to pick a winner.

Continue the decision