Do coding agents recommend Cloudflare Turnstile?
Cloudflare Turnstile is the most chosen bot protection service, taking 57% of 160 judged bot protection sessions. Measured with Claude Code, Codex and Cursor.
Cloudflare Turnstile is the most chosen bot protection service, taking 57% of 160 judged bot protection sessions. It was also raised as a candidate in 67 further sessions without being chosen.
This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in bot protection inside a realistic codebase. Not what a chat assistant says about Cloudflare Turnstile. What an agent actually installed.
The numbers
| Category | Bot protection |
| Sessions in the category | 160 |
| Sessions where Cloudflare Turnstile was chosen | 91 |
| Install share | 57% |
| Rank in category | 1 of 6 |
| Codebases it won in | 5 |
| Raised as a candidate, not chosen | 67 |
| Chosen when considered | 58% |
| Site | cloudflare.com |
By agent
The three agents treat Cloudflare Turnstile very differently. Codex chose it in 74% of its sessions in this category and Cursor in 38%, a spread of 36 points.
| Agent | Sessions | Chose Cloudflare Turnstile | Share |
|---|---|---|---|
| Claude Code | 54 | 31 | 57% |
| Codex | 54 | 40 | 74% |
| Cursor | 52 | 20 | 38% |
A single blended install share would report the midpoint and hide both ends. If most of your users run Cursor, your real position here is 38%, not 57%.
By who was asking
Cloudflare Turnstile does much better with one kind of buyer than another. It won 74% of sessions asked as vibe coder and 31% of those asked as enterprise team.
| Who is asking | Sessions | Chose Cloudflare Turnstile | Share |
|---|---|---|---|
| Vibe coder | 61 | 45 | 74% |
| Junior developer | 63 | 35 | 56% |
| Enterprise team | 36 | 11 | 31% |
What Cloudflare Turnstile was up against
The full ranking in bot protection, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Cloudflare Turnstile (this page) | 91 | 57% |
| 2 | Vercel BotID | 18 | 11% |
| 3 | Built in-house (no product adopted) | 18 | 11% |
| 4 | ALTCHA | 12 | 8% |
| 5 | django-axes | 11 | 7% |
| 6 | Google reCAPTCHA | 6 | 4% |
| 7 | Laravel RateLimiter | 2 | 1% |
What this means
Leading a category is a position to defend rather than a result to celebrate. In our data, leaders lose ground in exactly two situations: when the person asking is an enterprise buyer with procurement constraints, and when a competitor's documentation is easier for an agent to integrate correctly.
The defence is unglamorous. Keep the quickstart running. Keep the version current on the page. Keep the names aligned. Stay in the templates.
Where these numbers come from
The 160 sessions in bot protection are part of a published set of 5,292, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every bot protection run can be replayed on the board.
If you work on Cloudflare Turnstile: the judge recorded a reason for every session where it was raised and passed over. Those reasons are in the transcripts.
<!-- generated by scripts/write-data-pages.mjs -->
Common questions
Do coding agents recommend Cloudflare Turnstile?
Yes. Cloudflare Turnstile was chosen in 91 of the 160 judged sessions in bot protection, a 57% install share, ranking first in its category.
Does Claude Code recommend Cloudflare Turnstile?
In 31 of the 54 sessions in bot protection run with Claude Code, which is 57%.
Do different coding agents treat Cloudflare Turnstile differently?
Yes, and by a wide margin. Codex chose it in 74% of its runs and Cursor in 38%.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 51 realistic codebases and asked to solve real tasks. A simulated project owner approved or questioned each recommendation before any code was written, and a judge from a model family that builds none of the agents read every session blind.
How often is Cloudflare Turnstile considered but not chosen?
It was raised as a candidate in 67 sessions without being chosen, and chosen in 91. That is a 58% conversion from considered to chosen.
Where this comes from
Armature ran 5,292 judged sessions with Claude Code, Codex and Cursor inside 51 realistic codebases, and published every run. The numbers on this page come from that work.