UX Tools Survey · Spring 2026

State of Prototyping: Spring 2026

The most-used design tool after Figma is now an AI.
The profession split into three camps, and most people haven’t noticed yet.

We surveyed 1,478 designers and builders across 18 regions on how they actually work right now. What they use every week. How much they’re vibe coding. Whether they trust AI to ship. And what they’re investing in next.

This is an open data project. The full de-identified dataset is available under CC BY 4.0. You can download the raw data, query it through the open API, or connect it to your AI tools via the MCP server.

1. Who took this survey

State of Prototyping is part of our ongoing survey initiative to understand how design and software are changing. This Spring 2026 edition is the first snapshot, built for designers and builders who work across the design-to-code spectrum.

61.4% of respondents are outside North America. Western Europe (16.2%), South Asia (8.1%), and Southeast Asia (3.8%) are the largest non-NA regions. This is a global sample.

Where Designers Work

Company size and work setting · 1,478 respondents, Spring 2026

Startup (2–100)
29.3%
Startup (2–100)
29.3%
Independent / Freelance
17.9%
Independent / Freelance
17.9%
Enterprise (1,000+)
17.7%
Enterprise (1,000+)
17.7%
Mid-size (101–999)
15.4%
Mid-size (101–999)
15.4%
Agency / Consulting
13.5%
Agency / Consulting
13.5%
Student / Between
6.2%
Student / Between
6.2%

n = 1,478 · Spring 2026

A Global Sample, Not a North America-Only Read

Top regions in the sample · 61.4% of respondents are outside North America

North America
38.6%
North America
38.6%
Western Europe
16.2%
Western Europe
16.2%
South Asia
8.1%
South Asia
8.1%
Eastern Europe
7%
Eastern Europe
7%
Northern Europe
7%
Northern Europe
7%
Southeast Asia
3.8%
Southeast Asia
3.8%
Southern Europe
3.4%
Southern Europe
3.4%
South America
3.3%
South America
3.3%
Australia & New Zealand
3.2%
Australia & New Zealand
3.2%
Western Asia (Middle East)
2%
Western Asia (Middle East)
2%
West Africa
1.9%
West Africa
1.9%
Northern Africa
1.2%
Northern Africa
1.2%
Central Asia
1%
Central Asia
1%
East Asia
0.9%
East Asia
0.9%
East Africa
0.8%
East Africa
0.8%
Southern Africa
0.8%
Southern Africa
0.8%
Central America & Caribbean
0.6%
Central America & Caribbean
0.6%
Central Africa
0.3%
Central Africa
0.3%

n = 1,476 · Spring 2026

2. The stack right now

Five of the ten most-used weekly tools are now AI. That sentence did not make sense two years ago.

Figma holds its seat. But Claude, ChatGPT, Claude Code, Figma Make, and Gemini now sit alongside it in the weekly rotation. The traditional design tool stack has been restructured from the inside.

5 of the Top 10 Weekly Tools Are Now AI

What designers use every week · % of respondents

Figma
82.6%
Figma
82.6%
Claude
50.8%
Claude
50.8%
ChatGPT
48.2%
ChatGPT
48.2%
Claude Code
38.4%
Claude Code
38.4%
Figma Make
34.8%
Figma Make
34.8%
FigJam
34%
FigJam
34%
Slack
32.7%
Slack
32.7%
Gemini
32.3%
Gemini
32.3%
Google Meet
24.8%
Google Meet
24.8%
Notion
24.5%
Notion
24.5%

n = 1,478 · Spring 2026

The #2 Weekly Tool

Claude

Claude

Claude is the #2 weekly tool in design, after Figma. 50.8% of designers use it every week — more embedded in designer workflows than any canvas-first tool.

Editor’s note

An AI coding terminal is now more embedded in designer workflows than any canvas-first tool. That shift happened quietly, and it happened fast.

3. The vibe coding split

This was the centerpiece question: how much of your building time is actually vibe coding, using AI to generate code you may not fully understand, but that works?

The profession split into thirds.

Vibe Coding 50%+

43.8%

43.8%

of designers now spend more than half their building time on AI-generated code

The Profession Has Split Into Thirds

How much of your building time is AI-generated code?

None (0%)
37.7%
None (0%)
37.7%
Occasionally
18.5%
Occasionally
18.5%
About half
12.7%
About half
12.7%
Most of it
17.5%
Most of it
17.5%
Nearly all
13.6%
Nearly all
13.6%

n = 1,478 · Spring 2026

Editor’s note

37.7% of designers do zero vibe coding. 31.1% say it’s most or all of how they build. These are not different generations. They work in the same orgs, on the same products, in the same Slack channels.

43.8% spend more than half their time vibe coding. 31.1% say it’s most or nearly all of how they build. And 37.7% do zero.

The hype has outpaced adoption at the tail, but the more surprising number is the 37.7% who haven’t started. Not “a little.” Zero.

4. Vibe coding by role

Same profession. Different reality.

The vibe coding split is not random. It maps almost perfectly to role type.

An 80.9% vs 35.0% Split in the Same Design Org

% spending 50%+ of building time on AI-generated code, by role

Design Engineer
80.9%
Design Engineer
80.9%
Lead / Principal
56.8%
Lead / Principal
56.8%
Non-designer
50.9%
Non-designer
50.9%
Manager / Director
46.6%
Manager / Director
46.6%
IC Designer
35%
IC Designer
35%
Researcher
26.1%
Researcher
26.1%

n = 1,478 · Spring 2026

The managers-at-46.6% number is telling. Vibe coding didn’t just serve engineers. It gave managers and non-designers an exit from prototyping constraints they’ve always had.

5. Who’s building their own tools

More than half of designers are now builders. Not for their company. For themselves.

59.1% of designers have built their own tool, app, or utility with AI in the last 6 months. One in four does it regularly.

That number would have been in the low single digits two years ago. The ability to build something for yourself, not a product, just a thing that solves your own problem, has unlocked a new kind of creative leverage.

59.1% of Designers Have Built Their Own AI Tool

Have you built your own tool, app, or utility with AI? · last 6 months

Yes, once or twice
33.7%
Yes, once or twice
33.7%
No, but I want to
30.5%
No, but I want to
30.5%
Yes, I do it regularly
25.3%
Yes, I do it regularly
25.3%
No, don’t plan to
10.4%
No, don’t plan to
10.4%

n = 1,476 · Spring 2026

59.1% have built something. 30.5% want to but haven’t yet. Only 10.4% have no plans to.

The intent-to-try cohort is bigger than the never-will cohort by 3x. The builder instinct is spreading across the design org.

6. The trust line

Nearly 1 in 3 designers trust AI output for production with review. That is a bigger number than most people would have guessed.

Only 1.4% trust it without any oversight. But 32.8% trust it enough to ship, with a review pass. We are in the “first drafts I heavily edit” era. The trust line is not whether AI works. It is how much cleanup it takes before it is usable.

32.8% Trust AI for Production With Review

How far do you trust AI-generated output in your workflow?

First drafts, edit heavily
34.2%
First drafts, edit heavily
34.2%
Exploration only
29.2%
Exploration only
29.2%
Review before shipping
24.7%
Review before shipping
24.7%
Ships with minor tweaks
8.1%
Ships with minor tweaks
8.1%
Don’t use AI output
2.4%
Don’t use AI output
2.4%
Full trust, no oversight
1.4%
Full trust, no oversight
1.4%

n = 1,477 · Spring 2026

7. What’s blocking everyone

The #1 thing blocking designers from AI isn’t the quality of the output. It’s time.

The top 3 blockers are within 3.5 percentage points of each other: time to learn tools (55.7%), too many tools to evaluate (53.0%), and AI output quality (52.2%). That’s not noise. That’s signal. Three simultaneous problems, nearly identical in size.

The Top 3 Blockers Are Within 3.5 Points of Each Other

What’s slowing down your workflow the most?

Time to learn tools
55.7%
Time to learn tools
55.7%
Too many tools
53%
Too many tools
53%
AI output quality
52.2%
AI output quality
52.2%
Budget / procurement
34.2%
Budget / procurement
34.2%
Security / compliance
28.9%
Security / compliance
28.9%
Engineering constraints
19.4%
Engineering constraints
19.4%

n = 1,478 · Spring 2026

The industry built a hundred AI tools and forgot to give anyone time to learn them. The blocker isn’t capability. It’s bandwidth.

8. How workflows changed in 6 months

71.1% have added AI or gone AI-central in the last 6 months. Only 9.9% say “mostly the same.”

The shift already happened. The question is how far you’ve gone.

71.1% Have Added AI or Gone AI-Central in 6 Months

How has your design workflow changed since late 2025?

Added AI tools
36.5%
Added AI tools
36.5%
AI is now central
34.6%
AI is now central
34.6%
Still in flux
15.2%
Still in flux
15.2%
Mostly the same
9.9%
Mostly the same
9.9%
Consolidated tools
3.8%
Consolidated tools
3.8%

n = 1,478 · Spring 2026

Startups vs Enterprise: A 4.1-Point Gap

Share saying “AI is now central” by work context

Startup (2–100)
38.8%
Startup (2–100)
38.8%
Independent / Freelance
36.2%
Independent / Freelance
36.2%
Mid-size (101–999)
34.8%
Mid-size (101–999)
34.8%
Enterprise (1,000+)
34.7%
Enterprise (1,000+)
34.7%
Agency / Consulting
28.6%
Agency / Consulting
28.6%
Student / Between
21.7%
Student / Between
21.7%

n = 1,478 · Spring 2026

The gap is 4.1 points. Much narrower than most people expect.

9. How designers feel about their role

Your confidence tracks with your seat. And the gap between the most optimistic and most anxious groups is the sharpest finding in this survey.

Design Engineers Feel More Valuable. Researchers Feel Most at Risk.

How do you think AI will affect your role in the next 2 years?

Role

More Valuable

Less Secure

About Same

Design Engineer

50.0%

10.6%

10.6%

Lead / Principal

43.2%

20.0%

24.6%

Manager / Director

35.4%

23.6%

22.4%

Non-designer

31.6%

24.6%

24.6%

IC Designer

24.8%

32.4%

21.7%

Researcher

17.4%

39.1%

21.7%

n = 1,478 · Spring 2026

Design Engineer

50.0%

50.0%

feel more valuable

10.6%

10.6%

feel less secure

Researcher

17.4%

17.4%

feel more valuable

39.1%

39.1%

feel less secure

Same industry. Same wave of change. Opposite experience.

Editor’s note

IC designers and researchers are the most exposed. They are also the largest groups in most design orgs. If your org isn’t talking about this openly, it’s not because the problem isn’t there. It’s because no one wants to say it first.

10. Where designers are investing next

Two of the top 3 investment areas are AI. Design systems is the only holdout in the top tier, and that’s not an accident.

64.0% say AI-assisted coding is their top investment. Agent workflows at 46.3%. Design systems at 40.2%.

Systems work is how you stay relevant when the output layer gets automated. The designers investing in systems aren’t ignoring AI. They’re building the layer AI can’t replace yet.

2 of the Top 3 Investment Areas Are AI

Where are you investing your time in the next 12 months? (pick 3)

AI-generated coding
64%
AI-generated coding
64%
Agent workflows
46.3%
Agent workflows
46.3%
Design systems & tokens
40.2%
Design systems & tokens
40.2%
Canvas design tools
21.3%
Canvas design tools
21.3%
Video, motion & 3D
20.4%
Video, motion & 3D
20.4%
Simplifying my stack
17.5%
Simplifying my stack
17.5%
Image generation
14.7%
Image generation
14.7%
No-code (not AI)
13.7%
No-code (not AI)
13.7%
Manual coding
9.5%
Manual coding
9.5%
No major changes
6.1%
No major changes
6.1%

n = 1,478 · Spring 2026

11. The satisfaction gap

1.5 points separate the floor from the ceiling. Both are people with the same job title. The difference is how much of their workflow runs on AI.

No vibe coding: 5.93/10. Heavy vibe coders: 7.39/10. The satisfaction gradient is nearly linear across adoption levels.

Heavier Vibe Coders Are More Satisfied

Mean workflow satisfaction (1–10) by vibe coding level

None (0%)
5.93
None (0%)
5.93
Occasionally
6.12
Occasionally
6.12
About half
6.83
About half
6.83
Most of it
7.16
Most of it
7.16
Nearly all
7.39
Nearly all
7.39

n = 1,477 · Spring 2026

The Satisfaction Gap

6.49/10

6.49/10

Overall mean workflow satisfaction. ▲ +1.46 heavy vibers vs. non-vibers. None: 5.93 → Nearly all: 7.39.

This is not proof that vibe coding causes satisfaction. Cross-sectional surveys can’t prove causation. People who adopt new tools faster may also be more satisfied by temperament. But the correlation is strong enough, and linear enough, that dismissing it requires an explanation too.

12. The story in summary

Six findings that capture what this data actually says.

01

Claude is the #2 weekly tool in design, after Figma.

50.8% of designers use it every week. Claude Code sits at #4, ahead of FigJam. An AI terminal outranks every canvas-first and collaboration tool in the stack.

02

The profession split into thirds, and they barely recognize each other.

No vibe coding: 37.7%. Some: 18.5%. Majority AI-generated code: 43.8%. Design engineers at 80.9% vs. IC designers at 35.0%.

03

More than half of designers are building for themselves.

59.1% have built their own tool with AI in the last 6 months. One in four does it regularly. Only 10.4% have no plans to.

04

The biggest blocker isn’t AI quality. It’s time.

Time to learn tools (55.7%), Too many tools (53.0%), and AI output quality (52.2%) are within 3.5 points. The industry shipped a hundred tools and forgot to give anyone time to learn them.

05

Design engineers feel more valuable. Researchers feel most at risk.

Design Engineer: 50.0% more valuable, 10.6% less secure. Researcher: 17.4% more valuable, 39.1% less secure. Same industry. Opposite experience.

06

Vibe coders are measurably more satisfied.

No vibe coding: 5.93/10. Heavy vibe coders: 7.39/10. The gradient is nearly linear. Correlation, not causation, but strong enough to take seriously.

If this is you

If you’re an IC designer:

35.0% of your peers are majority vibe coding, which means 65% aren’t. You’re not behind. But the gap between you and design engineers is the widest split in the data. The question is whether you want to close it.

If you’re a manager:

46.6% of managers are vibe coding. You’re not watching from the sidelines anymore. The blocker data says your team needs time, not more tools. Protect their learning bandwidth.

If you’re a researcher:

This data is hard to read. 39.1% of researchers feel less secure. That’s the highest anxiety in any role. But 17.4% feel more valuable, which means the path exists. The researchers who are thriving are the ones integrating AI into their practice, not waiting for permission.

If you’re a design engineer:

You’re in the most optimistic seat in the industry right now. 80.9% of your peers are majority vibe coding. Your investment in code + AI fluency is paying off. The data says: keep going.

Methodology

The State of Prototyping Spring 2026 survey ran from March 14 to April 6, 2026 and collected 1,478 responses from designers and builders across 18 world regions. Distribution was via UX Tools newsletter, social channels, and sponsor networks.

Published charts use aggregated data, and the full de-identified microdata is also available for download. One malformed role response in the raw CSV is excluded from role-based breakdown tables. The “Researcher” role (n=23) should be treated as directional only due to small sample size. Region breakdowns use n=1,476 because two respondents did not provide a region. Multi-select questions sum to more than 100%. Cross-tab percentages are calculated within each role’s n.

The full dataset is available under CC BY 4.0. Citation: UX Tools. (2026). State of Prototyping Spring 2026. https://survey.uxtools.co.

Have a question about the methodology? Want to run your own analysis? The full dataset and API are open. Prove us wrong.

State of Prototyping · Open Survey + Open Data

An open survey project by Tommy Geoco and UX Tools. Spring 2026 includes n = 1,478 responses, a public dataset and a public API. Licensed under CC BY 4.0.