Compare · 2 or 3 programs

Put them next to each other.

Same fields, no narrative padding. Share the URL. It already holds the selection. Visa prior is a planning range, not a score. Indices, future proof, job security, all-in sketch, holds, costs you, and full freight are defined in the word list under a finished comparison.

Three is the maximum. Remove one to swap.

Berkeley, California

UC Berkeley MIMS

2 years

Livability

5 / 10

Livability: 5 out of 10

Food, transit, weather, and the intellectual scene are excellent. Housing is the crisis, and neighborhood safety is uneven. The school’s own housing line is $19,920 a year, which means roommates and compromises. Livability is high only if that budget does not scare you.

Job security

9 / 10

Job security: 9 out of 10

Best published salary and the clearest PM network. UX and product design outcomes are real and secondary. Job security here means ‘product roles in the Bay,’ not ‘any design studio.’

Future proof

9 / 10

Future proof: 9 out of 10

AI, data, policy, and human-centered work in one school is the most systems-heavy mix in the set. You can leave under-practiced in visual craft. You are unlikely to leave unable to talk about how AI products are built and governed.

Selectivity

8 / 10

Selectivity: 8 out of 10

No rate printed. A 50–70 person class with a 3.57 GPA and five-year experience average is a tight door. Holistic review helps nontraditional paths. It does not lower the standard.

ROI

6 / 10

ROI: 6 out of 10

The salary data is the best in the set and the two-year cost is tied for worst. Fellowships are too small and too uncertain to repair the sticker. Worth it for a PM or technical product outcome. A poor bet for a generalist design year.

Visa prior 65–80%. STEM-designated, in the densest market of employers that already sponsor, with a curriculum that points at PM and engineering-adjacent titles. Those titles sponsor more often than visual design. The top of the range still assumes a competent search and a lottery. It is not an admission outcome.

Official program site

Pittsburgh, Pennsylvania

Carnegie Mellon MHCI

12 months · 3 semesters

Livability

7 / 10

Livability: 7 out of 10

Pittsburgh is the affordable, winter city in this set. Campus life is dense, rents are sane next to Seattle or Berkeley, and the cultural scene is real if smaller. You will not stumble into a design job by living here.

Job security

9 / 10

Job security: 9 out of 10

Strongest brand in the six for competitive PD, UXR, and PM interviews. Capstone and alumni network do actual work. Not a 10, because sponsorship and a one-year clock still drop people.

Future proof

9 / 10

Future proof: 9 out of 10

HCII’s identity is human–AI interaction plus professional practice. Electives across CMU let you add technical depth if you move quickly. The risk is breadth squeezed by the capstone, not an outdated syllabus.

Selectivity

8 / 10

Selectivity: 8 out of 10

Only published rate in this set, at 23% for 2025–26. Cohort is about 57. The bar is the quality of the pool and the 500-word plus video filter, not a single-digit admit rate.

ROI

7 / 10

ROI: 7 out of 10

You pay roughly $93k in tuition with no aid, then a modest Pittsburgh year. One year of lost salary and a strong signal is a fair trade if you can write the check. It is a poor trade on high-interest debt.

Visa prior 60–75%. STEM-designated, which matters for the 24-month STEM OPT extension. The brand opens doors at employers that have sponsored before. Design studios often will not. Capstone sponsors are not visa sponsors. International searches in the program’s own commentary run longer. This range is a planning prior, not a rate.

Official program site

Seattle, Washington

UW MHCI+D

1 year · 46 credits minimum

Livability

6 / 10

Livability: 6 out of 10

Seattle’s daily life is excellent if you like weather, water, and transit in the core. Rent is the tax. A score of 6 is the city with the cost dragged out of an 8.

Job security

8 / 10

Job security: 8 out of 10

Excellent for product design and UX where Seattle hires. Less automatic for PM than Berkeley, less nationally mythic than CMU, more craft-legible than either. The missing internship keeps it off a 9.

Future proof

7 / 10

Future proof: 7 out of 10

Prototyping and human-centered practice age well, including for AI products, if you point projects at them. A one-year studio will not give you Berkeley’s systems depth or Georgia Tech’s computing track. Electives are real and few.

Selectivity

7 / 10

Selectivity: 7 out of 10

Small cohort, required video, portfolio that has to show making. Harder than the large information-school master’s programs. Not a published-rate story.

ROI

7 / 10

ROI: 7 out of 10

Tuition near $64k and a high-rent year, then you earn. Better cash shape than Berkeley or Michigan. Worse than Austin on tuition, better if craft was the thing holding your career back.

Visa prior 55–70%. The apply page calls the program a STEM program. Seattle employers sponsor, and many design roles still do not. No internship means fewer conversion paths. The top of the range is a strong portfolio aimed at large product orgs. The bottom is a studio search that starts in May.

Official program site

City names are places, not scores. Updated 11 October 2026. Visa ranges are editorial planning priors, not school statistics. No program on this list publishes an OPT-to-sponsorship rate. The range assumes a STEM-eligible I-20, a portfolio aimed at companies that have sponsored before, and a 2026 hiring market in which design studios sponsor less often than large tech and some product orgs. H-1B remains a lottery. Update your plan when policy or the market moves.

Word list
Decision framework
The rules for reading the order on the homepage. It is a weighting for one applicant, not a quality ladder of the six schools.
Indices
Five scores from 1 to 10: livability, job security, future proof, selectivity, and ROI. Each score compares these six programs only.
Visa prior
A planning range for OPT and for whether employers in that market sponsor. It is not an admission rate and not a promise.
Future proof
How directly the curriculum treats human-AI work, systems, and the mix of research, craft, and technical skill after 2026.
Job security
How usable the path is into product design, UX research, or product management: brand, network, internship, and where alumni land.
All-in sketch
Tuition plus a living estimate for the length of the degree. Not an invoice. Fees, flights, and aid sit outside it unless a line says so.
Holds
Where the program is strong for this decision.
Costs you
What you give up, or the risk you take, if you pick it.
Full freight
You pay the published tuition. Do not assume a scholarship or an assistantship will cut it.