Findings / run 2026-11-02
Radix is everywhere in the answers, and nowhere in the install commands.
In 30 AI coding answers that needed UI components, shadcn/ui was installed 19 times. Radix UI was mentioned in 11 answers and installed directly in none.
| Product | Installed | Claude | GPT | Gemini | Mentioned |
|---|---|---|---|---|---|
| shadcn/ui | 19 | 7 | 6 | 6 | 19 |
| MUI | 3 | 0 | 2 | 1 | 3 |
| Ant Design | 1 | 0 | 0 | 1 | 1 |
| Radix UI (primitives) | 0 | 0 | 0 | 0 | 11 |
| Mantine | 0 | 0 | 0 | 0 | 1 |
Five tasks that need a React component library (dialogs, dropdowns, forms, dark mode), twice on each of Claude, GPT and Gemini. shadcn/ui was installed in 19 of 30 answers. 7 answers hand-rolled components with Tailwind, MUI got 3 and Ant Design 1.
Radix UI was mentioned in 11 answers and installed directly in none. It always appears as the thing underneath:
“Use shadcn/ui (Radix primitives + Tailwind). You own the component code…” Claude
“…highly polished, deeply accessible components built on Tailwind CSS and Radix UI…” Gemini
This one is less of a loss than it looks: shadcn/ui components pull in @radix-ui/* packages, so Radix still ends up in the lockfile. But the developer's decision, and the brand they remember, is shadcn/ui.
What it means
A mention isn't a pick, and a dependency isn't a choice. For a library that lives underneath another one, the number that matters is who the model names in its first line. Here that's shadcn/ui on all three models (Claude 7, GPT 6, Gemini 6).
Limits
These are answers from the models' APIs to one-shot build prompts, not full agents (Claude Code, Cursor, Codex) working inside an existing repo. Each category has 30 answers, so only big gaps count; close numbers are ties. How we measure, and every answer is in the raw data.
See the data
- UI component library: all prompts and answers
- Headless CMS: all prompts and answers
- E2E testing: all prompts and answers
More findings
- Supabase is mentioned, never installed: the Postgres host AI agents pick
- Drizzle vs Prisma: which ORM Claude, GPT and Gemini install
- Sentry and Playwright: the only tools AI agents install every time
- Which email API do AI agents install? Resend 16, Postmark 4, Mailgun 0
- Datadog was named 9 times by AI coding agents and installed 0 times
- Cloudflare R2 vs S3: AI agents name R2, then install the AWS SDK
- pgvector beats Pinecone, Qdrant and Chroma in AI-written RAG code
- Where AI coding agents deploy: Vercel 12, Netlify, Railway, Render and Fly.io 0
Did we get something wrong?
A tool we missed, a pick that looks wrong, a prompt we should add? Tell us. We read every note and fix the data in the next run.
Is your product losing these picks?
Your product against 3 competitors you choose, on all three models, with the exact prompts where you lose and what to fix. $19, delivered within 24 hours.