ML Engineer Salary in Washington StateSeattle, Bellevue, Redmond. No state income tax. RSU-heavy.

Washington's ML labour market is concentrated almost entirely in the Greater Seattle area (Seattle proper, Bellevue across Lake Washington, and Redmond to the east). The three sub-metros are functionally a single market for ML hiring: cross-employer mobility within the Eastside / Seattle corridor is high, and salary bands are similar to within 5 percent. Eastern Washington (Spokane and the Tri-Cities) has a much smaller ML market with materially lower bands, mostly anchored by remote roles for Seattle-area employers or PNNL (Pacific Northwest National Laboratory, federal pay scale). The intra-state spread for equivalent level is roughly 50 to 60 percent between Bellevue and Spokane.

Figure wa-1. Washington metro base salary bands and L5 total compensation, May 2026. Synthesised from Levels.fyi Seattle metro filter and BLS OEWS Washington 15-2051. Total comp assumes typical T2 hyperscaler RSU mix (40 to 50 percent of TC).

Washington is one of nine US states with no individual income tax on W2 wages. The structural after-tax advantage is the single most important reason to consider Seattle over the Bay Area for ML compensation that is heavily weighted toward base salary rather than equity. The advantage compounds over a career: $18,000 to $25,000 per year in additional take-home, invested at 7 percent, becomes approximately $250,000 to $350,000 in additional wealth over 10 years. For a senior ML engineer making the decision between Seattle and the Bay Area for a 4 to 8 year career stage, this is a non-trivial input.

The caveat is the 2022-introduced Washington capital gains tax: 7 percent on long-term capital gains above $250,000 per year, currently tested in court but operational. This applies to gains on shares sold, not to RSU vesting (which is taxed as wage income, not capital gains). For most ML engineers, the practical impact is limited; only large equity-realisation events (founder cash-out, large pre-IPO stock sales) trigger the tax. The lower threshold of $250,000 makes this more punitive than many comparable state-level capital-gains taxes elsewhere.

Figure wa-2. Hypothetical take-home cash on $250,000 base salary by state, single filer, 2026. Excludes equity, 401(k), HSA. Simplified federal calculation. Source: state tax department schedules.

Seattle ML compensation is more RSU-weighted than any other major US metro. Both anchor hyperscalers headquartered in the area run a predictable model: initial 4-year RSU grant at hire, then annual performance-and-stock-price-sensitive refresh grants. A typical L5 ML engineer offer in 2026 includes initial RSUs of $300,000 to $500,000 (vesting back-loaded over 4 years), plus annual refresh grants of $100,000 to $200,000 starting in year 2. At public-company stable share prices, this is straightforward realised cash; at declining share prices, the refresh-grant model effectively rebases a candidate's effective compensation each year.

The arms race for AI talent has compressed the year-over-year refresh cadence at both anchor employers since 2022. A senior ML engineer who would have received $80,000 to $100,000 in annual refresh in 2020 might now receive $150,000 to $250,000 if they hold scarce foundation-model or LLM-serving expertise. The retention dynamic is legible to candidates: bring a competing offer from a Bay Area T1 lab and the Seattle employer will typically respond with an accelerated refresh grant rather than risk losing the headcount.

The risk is correlation: an ML engineer holding $1,000,000+ of unvested RSUs from a single employer carries concentration risk that is higher than the same person's nominal job security would imply. Stock-price declines simultaneously reduce realised compensation, reduce the value of the unvested grant, and reduce the employer's willingness to issue refresh grants. Bay Area peers at T1 frontier labs hold even more concentrated risk (pre-IPO equity is illiquid), but with potentially larger upside.

A practical heuristic: target unvested RSU concentration below 200 percent of annual base salary, and sell RSUs at vest to diversify (paying ordinary income tax on the vesting event, no incremental capital gain). Seattle's lack of state income tax means selling at vest is less painful than in California, where the same diversification decision generates a 9.3 percent state tax cost.

The BLS OEWS Washington state page reports SOC 15-2051 (Data Scientists) annual mean wage above $160,000 (2024 vintage), with the Seattle / Bellevue / Tacoma MSA ranking in the top three metros nationally. SOC 15-1252 (Software Developers) shows a similar pattern, with the Seattle metro top three for both mean and 90th-percentile wages.

A caveat that recurs in tech-comp analysis: BLS OEWS captures W2 wages, which lags actual market compensation because the data collection cycle is annual and excludes equity vesting beyond the most recently reported tax year. ML engineer compensation has been inflating at 8 to 15 percent per year in Seattle for the senior and staff levels since 2022, and OEWS is structurally late to reflect this. For real-time comp data, Levels.fyi self-reports filtered by Seattle metro give a better lens, with the self-selection caveat that respondents skew higher-comp.

Washington also publishes wage data through the Employment Security Department Labor Market Information portal. Both ESD and BLS show the Seattle metro at the top of the state distribution; Eastern Washington and the Olympic peninsula have materially lower bands. The Bellingham / Whatcom County region has limited ML employment despite proximity to the Vancouver, BC tech ecosystem.

1. BLS Occupational Employment and Wage Statistics, Washington, 2024
2. Levels.fyi Software Engineer, Seattle metro area
3. Washington Employment Security Department Labor Market Information
4. Washington State Department of Revenue (no state income tax confirmation)
5. BEA Regional Price Parities by metro