Amazon Engineering Levels and Salary: L4 to L10, and Why Year One Pay Is Not What You Think

No, and this is the most persistently out of date fact circulating about Amazon compensation. Amazon raised its United States base salary cap from $160,000 to $350,000 in 2022, specifically because the old cap was costing it senior candidates against Google and Meta.

In practice most engineers do not approach the new ceiling. Base salary at L4 sits around $141k and at L6 around $210k. The cap matters most at L7 and above, where the old limit used to force almost the entire package into stock.

It is how Amazon vests stock grants across four years: five per cent in year one, fifteen per cent in year two, then forty per cent in each of years three and four. Eighty per cent of your equity arrives in the second half of the grant.

Almost every other large technology company vests evenly or front loads. Amazon does the opposite, and it does it deliberately as a retention mechanism. Leaving at the two year mark means walking away from four fifths of the equity you were promised.

Because without them the first two years of an Amazon offer would look uncompetitive. With only twenty per cent of the grant vesting across years one and two, the sign on bonuses exist to fill that hole and make the offer comparable at the point where you are deciding.

The important thing to understand is that they stop. Your year three compensation contains no sign on bonus at all, which is fine because that is when the large vests begin. The risk is treating the year one number as your steady state salary, because it is not, and it is the number recruiters quote.

Not in any meaningful sense for engineers. The engineering ladder runs L4, L5, L6, L7, L8, and then L10. L9 exists on paper as something close to a placeholder and is not a rung people are promoted into.

This trips up cross company comparisons, because guides that assume a continuous numbering sequence end up mapping Amazon levels one rung off. L10 Distinguished Engineer is the top of the individual contributor ladder.

No, and assuming so is an expensive mistake. Amazon L6 maps roughly to Google L5 and Meta E5. The higher number does not mean higher seniority, because the two ladders simply start counting in different places.

Amazon starts its corporate ladder at L4 while Google starts engineering at L3, which accounts for most of the offset. When a recruiter cites your level as evidence for an offer, check the mapping rather than the number.

As of 2026 Amazon offers eligible employees an annual election to take a portion of the restricted stock unit grant as cash instead, paid out on the same schedule the stock would have vested on.

It reduces your exposure to the share price, which is a genuine benefit if your finances cannot absorb that volatility, and it caps your upside if the stock performs. It does not change the back loaded timing, so it solves the concentration problem rather than the cash flow problem.

L6 to L7 in absolute terms. The median moves from roughly $413k to roughly $732k, an increase of more than $300k, which is larger than the entire L4 package.

In proportional terms L5 to L6 is comparable and far more achievable. It is also the promotion most engineers can realistically plan for, since L7 depends on organisational scope existing rather than on your performance alone.

Badly in years one and two, better in years three and four. Google front loads its vesting at roughly thirty three per cent in the first year, so a Google offer pays most in the period where an Amazon offer pays least.

The only sensible way to compare them is to build a four year table for each and look at the cumulative total, then discount for the risk that you leave before year three. If you have any real doubt about staying three years, the Amazon offer is worth materially less than its headline suggests.

A Bar Raiser is a trained interviewer from outside the hiring team who sits on the loop to protect hiring standards. They hold a veto, and they use it.

The role exists to stop hiring managers lowering the bar to fill a role quickly. In practice it means one of your interviewers has no stake in whether the position gets filled, which changes how that conversation goes. The interview process guide covers how the loop is structured.

No. Refreshers are granted annually based on performance and are the mechanism that keeps your compensation from falling off once the initial grant is exhausted. They are not contractual and their size varies considerably.

This matters more at Amazon than elsewhere because the initial grant is back loaded. By the time you are five years in, refreshers rather than your joining grant determine almost all of your compensation, and no part of that was promised at the point you signed.

No. Amazon does not publish its internal level structure, its promotion rubric, or its compensation bands. United States job postings do carry legally required pay ranges in several states, which is the only official data available and covers base salary only.

Everything else here comes from crowdsourced compensation reporting and community accounts. It is good enough to plan a negotiation around and specific enough to be wrong in an individual case.

Financially, probably not, and this is the clearest practical conclusion on this page. Leaving before the year three vesting cliff means collecting twenty per cent of your equity while the sign on bonuses that made years one and two look reasonable have already been paid and spent.

The structure is designed to make exactly that outcome unattractive. If you are confident you want three or more years, the back loading works in your favour and the later years are genuinely competitive. If you are not, an offer that vests evenly is worth more to you at the same headline number.