Software engineer salary calculator

§Software engineer compensation in 2026: pay is up for survivors, cuts are accelerating

The headline number is steady. Software developers in the US earn a median of $133,080 per year ($64 an hour, exactly matching a 2,080-hour work year), per BLS Occupational Employment Statistics for May 2024 (BLS OOH).

The 90th percentile sits at $211,450, the floor for top-decile work.

BLS projects the occupation to grow 17.9% from 2023 to 2033, adding roughly 304,000 jobs, second-largest gain of any US occupation (BLS Employment Projections).

The median itself barely moved through the 2022-2023 freeze. If you stop reading there, nothing happened.

But the median misses where the action is. US software compensation has bifurcated across four tiers that pay very differently for the same nominal job title. At the top tier, total comp for a senior engineer roughly doubled relative to 2018. At the bottom tier, comp growth tracks inflation and equity is a lottery ticket priced at face value.

And the cuts didn't end. Q1 2026 saw 81,747 tech layoffs, already 45-55% of all of 2025's total (Crunchbase News tracker). Amazon cut roughly 30,000 jobs in the last five months. Meta announced another 10% workforce reduction in April. Oracle's single-event 30K cut is the biggest of the year.

This round isn't from over-hiring like 2022. It's funded by the $725B that Amazon, Microsoft, Alphabet, and Meta plan to spend on AI infrastructure in 2026, a 77% YoY increase (CNBC). Headcount is being traded for GPUs.

The problem isn't the offer. It's getting to the offer.

§The four-tier reality of US software comp

The same "Senior Software Engineer" title can pay $180K total at a non-tech enterprise and $478K total at FAANG. Both are accurate. Glassdoor and Indeed averaging across them produces a number that describes nobody.

The calculator splits the market into four tiers. Here's what each looks like at the senior level (representative example, not the only data point):

FAANG / Elite (Google, Meta, Apple, Amazon, Microsoft, Netflix, plus Stripe and Databricks for the comp ceiling). Senior median total comp $400-650K. Equity is roughly half the package, delivered as RSU vesting over four years. Base alone is typically $200-230K. Detailed per-company breakdown lives at the FAANG salary calculator.

Big Tech (Snowflake, Airbnb, Pinterest, Cloudflare, Datadog, and most well-funded public infrastructure companies). Senior median total comp $280-420K, roughly 70-85% of FAANG. Equity still meaningful, base similar, bonus rate around 12% of base.

Mid-size (Series C-D private companies, established public companies that aren't tech-first, regional tech employers). Senior median total comp $200-280K, roughly 50-65% of FAANG. Equity is smaller (often 10-20% of TC), base is the dominant component. Bonus rate around 8% of base.

Startup (seed through Series B). Senior cash comp $130-180K, roughly 35-45% of FAANG cash. Equity is theoretical: the headline package looks larger ($250-400K nominal TC including equity), but the equity is paper at the most recent round price. Empirical exit data suggests 80%+ of seed and Series A startups exit at $0 (CB Insights startup graveyard). The realistic expected value of startup equity is the slider input on the right side of the calculator.

Why this matters for offers. Public salary listings (Glassdoor, Indeed, PayScale) typically report base only. A "senior software engineer earns $180K" headline at a top-tier employer can understate real comp by $200K or more. The calculator above uses TC as the primary number for exactly this reason: base alone is misleading at any tier that grants meaningful equity.

§The bar moved sideways. Some roles are gone entirely.

The common take on 2023 layoffs was "tech doesn't pay anymore." It still pays. The 2026 reality is more nuanced: pay is up for surviving roles, and several role categories are being structurally eliminated.

Cut at industrial scale. Customer support, QA testing, content moderation, junior code review, middle management. These were first to face AI substitution because their output is verifiable and routinizable.

In shortage. ML engineers, AI safety researchers, data infrastructure engineers, GPU and cluster operators. Roughly 275,000 AI-related job postings were open in the US simultaneously with Q1 2026's record cuts.

Bar is higher for surviving categories. Companies that used to fast-track new grads now require more system-design rounds for the same level. A 2026 senior-level loop at top-tier employers typically includes 2 coding rounds, 2 system design rounds, and a behavioral. In 2021 the same level often got hired with one system design pass.

AI screens resumes. Most large tech ATSes now run resume content through an LLM filter before any human sees it. Generic "senior software engineer" titles with no measurable impact get filtered out in batch.

Net effect: bifurcation. For experienced engineers in non-substituted specializations, this is a buyer's market for talent and a seller's market for individual offers.

For new grads, and for engineers in routinizable roles, the floor moved down. The lottery odds got worse without the prize going down.

§The AI premium is bigger than people think, and growing

The conventional number is "AI engineers earn 15-30% more." That number is two years stale.

PwC's 2025 AI Jobs Barometer measured the wage premium for roles requiring AI skills at 56%, up from 25% a year earlier (PwC).

Signify Technology's 2025-2026 ML benchmarks put senior ML engineer base at $160-190K with top-tier total comp reaching $350-550K+ (Signify).

Two caveats on the 56% headline:

• The premium is concentrated at senior+ levels. Early-career AI roles pay only 5-15% above general SWE comp; the 56% figure averages across seniority.
• "AI skills" as PwC measures it is broader than ML engineering. It includes prompt engineering, RAG implementation, LLM integration.

Where the real ceiling is. Frontier-lab engineers at OpenAI, Anthropic, and Google DeepMind earn substantially more. Total comp in the $700K-$2M range is normal at the staff level.

The premium reflects a supply problem. The pool of engineers who can ship LLM-integrated features grew slowly while every company in the S&P 500 started hiring for it. The supply problem won't resolve in 2026.

Practical implication. If you're a general backend or frontend engineer, the 12-month payback on building AI/ML skills (prompts, embeddings, evals, fine-tuning) is real and measurable. The 5-year payback on becoming a pure ML researcher requires a PhD or equivalent and is a different career.

§How to read the calculator above

The calculator runs four selectors and one conditional slider, then produces a year-1 TC breakdown plus a five-year cumulative cash projection.

Selectors:

• Level. Years of independent ownership, not chronological. Senior means you own systems and mentor juniors, not just six years on a resume.
• Company tier. Startup / Mid-size / Big Tech / FAANG. The tier determines bonus rate and equity weight: startups model at 5% bonus, mid-size 8%, big tech 12%, FAANG 18%.
• Location. Geographic multiplier (see below).
• Specialization. AI/ML adds 20%. DevOps and SRE add 12%. Web/frontend subtracts 5% (oversupplied). Backend and mobile are baseline.

What you see in the result:

• Hero. Year-1 total comp at your intersection. Risk-adjusted for non-FAANG tiers.
• Stats. Base range, year-1 equity vest, gross hourly from base.
• Pie. Year-1 comp mix (base / equity / bonus).
• Line. 5-year cumulative cash earned (probability-adjusted equity + base + bonus per year).
• Bar 1. Your level across all 4 company tiers.
• Bar 2. Career progression at your tier across all 5 levels.

§Geographic adjustment

Public comp surveys aggregate across the US. Real offers compress by location. The calculator covers 18 metro areas plus fully Remote, calibrated against BLS OEWS May 2024 area data for Software Developers and Levels.fyi cross-location FAANG samples.

Anchor data points from BLS May 2024 medians:

• San Jose-Sunnyvale-Santa Clara: $208,270 (Bay baseline).
• Seattle-Tacoma-Bellevue: $169,340 (0.81 vs Bay raw).
• New York-Newark-Jersey City: $161,970 (0.78 vs Bay raw).
• Boston-Cambridge-Newton: $154,240 (0.74 vs Bay raw).
• Austin-Round Rock-San Marcos: $133,070 (0.64 vs Bay raw).
• Chicago-Naperville-Elgin: $130,030 (0.62 vs Bay raw).

The multipliers shown in the dropdown run 0.05-0.10 above the pure BLS ratio at FAANG-anchored metros (Bay, Seattle, NYC, Boston) because Levels.fyi tech-tier comp at those locations is higher than BLS aggregate. At metros without significant FAANG offices (Atlanta, Dallas, Phoenix), the multiplier hugs the BLS aggregate more closely.

Representative bands in the calculator:

• Bay Area at baseline (1.00). Highest concentration of tech-tier comp data.
• Seattle, NYC, Boston, LA sit within 7-12% of Bay (FAANG anchors lift them above raw BLS).
• DC Metro, San Diego, Portland sit 15-20% below.
• Austin, Atlanta, Dallas, Denver, Chicago, Raleigh-Durham, Salt Lake City cluster at 23-30% below (BLS-aggregate dominant).
• Phoenix, Minneapolis, Pittsburgh at 30-32% below.
• Remote (US) averages 20% below baseline. Specific employer policies vary: Meta and Google apply roughly 10% haircuts on remote; Amazon historically doesn't haircut; smaller companies range widely.

The discount applies to base, equity, and bonus uniformly. The cross-tier bar chart stays at the Bay baseline so visual comparison between tiers is not distorted by your selected location.

§Equity at face value is a lottery ticket: adjust for it

This is the most-missed mistake in offer evaluation. Equity at startups, late-stage private companies, and even some big tech is quoted at the most recent round price. That number only converts to cash if a liquidity event occurs at or above that valuation.

The slider on the right surfaces a realistic probability:

• Startup (default 15%). Empirical exit data: roughly 80% of seed and Series A startups exit at $0 (CB Insights aggregate). 15% reflects the long-run expected value of equity from a typical pre-Series-B company. Move higher if your specific company has clearer traction.
• Mid-size (default 60%). Late-stage private or recently public. Secondary markets exist, IPO path is more credible. Adjust based on the company's actual stage.
• Big Tech (default 90%). Publicly listed RSUs with normal share-price risk during the four-year vest.
• FAANG (slider hidden, fixed at 100%). Deeply liquid public stock. Stock price risk only.

When you change the slider, the hero, the pie chart, the equity-Y1 stat, and the five-year cumulative line all switch to the risk-adjusted number. The cross-tier bar chart stays nominal so comparison against peers is apples-to-apples.

§The five-year cumulative cash chart

This is the killer comparison metric for "FAANG vs startup" decisions. Single-year TC headlines flatter early-stage equity-heavy offers. Five-year cumulative cash exposes which offer actually pays more in real dollars.

The math: base + bonus is paid every year (constant). Equity vests at 25% per year for years 1-4, then drops to zero in year 5 because the initial grant is exhausted (refresher mechanics are FAANG-specific and modeled in the dedicated FAANG salary calculator). Each year's equity slice is multiplied by your probability slider before adding to the cumulative total.

The chart shows cumulative cash earned by year-end of years 1 through 5, with year-4 marked as the reference line because that's when the original equity grant fully vests. After year 4, the line flattens at startups (no equity refresher), and continues climbing at FAANG (where refreshers sustain TC, modeled on the FAANG page).

The most useful read: pick a FAANG senior offer and a startup senior offer with similar headline TC. Set the startup probability slider to 15%. Compare the year-5 cumulative number. Most of the time, the FAANG cumulative is roughly double the startup. That gap is what most engineers never compute before deciding.

§Compare two paths

The button above the calculator switches it into a side-by-side view: Path A on the left, Path B on the right. Each has its own complete set of selectors, slider, and results.

This is built for the exact decision the calculator is for. Examples:

• FAANG senior vs startup senior at same headline TC. Set both to Senior, FAANG vs Startup. The cumulative line chart shows the real-dollar gap after probability adjustment.
• Bay Area vs Austin at the same tier and level. The hero numbers diverge by roughly 20%. Useful when weighing a relocation.
• AI/ML versus backend at the same tier and level. The 20% specialization multiplier compounds over five years into a meaningful cumulative gap.

The bar charts in each panel still compare against the full tier set at that panel's level, so cross-tier context is preserved while you compare two specific paths.

The "Base / hour" stat is gross hourly from base salary only, computed against 2,080 hours per year. It's useful for comparing salaried roles to contract rates, not for calculating take-home pay.

For the dedicated FAANG per-company breakdown (Google L5 $425K, Meta E5 $478K, Stripe L4 $766K, Databricks L5 $647K, and more), with year-by-year refresher math and the private-company stock realization slider, see the FAANG software engineer salary calculator.