Do coding agents recommend Cloudflare Workers?
Cloudflare Workers was chosen in 7% of 287 judged serverless compute sessions, ranking fourth. Measured with Claude Code, Codex and Cursor.
Cloudflare Workers was chosen in 7% of 287 judged serverless compute sessions, ranking fourth. It was also raised as a candidate in 115 further sessions without being chosen.
This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in serverless compute inside a realistic codebase. Not what a chat assistant says about Cloudflare Workers. What an agent actually installed.
The numbers
| Category | Serverless compute |
| Sessions in the category | 287 |
| Sessions where Cloudflare Workers was chosen | 20 |
| Install share | 7% |
| Rank in category | 4 of 13 |
| Codebases it won in | 3 |
| Raised as a candidate, not chosen | 115 |
| Chosen when considered | 15% |
| Site | workers.cloudflare.com |
By agent
The three agents land within 6 points of each other on Cloudflare Workers, which is closer than most products in this experiment manage.
| Agent | Sessions | Chose Cloudflare Workers | Share |
|---|---|---|---|
| Claude Code | 96 | 3 | 3% |
| Codex | 96 | 9 | 9% |
| Cursor | 95 | 8 | 8% |
By who was asking
Cloudflare Workers does much better with one kind of buyer than another. It won 42% of sessions asked as vibe coder and 0% of those asked as enterprise team.
| Who is asking | Sessions | Chose Cloudflare Workers | Share |
|---|---|---|---|
| Vibe coder | 36 | 15 | 42% |
| Junior developer | 72 | 3 | 4% |
| Senior engineer | 107 | 2 | 2% |
| Enterprise team | 72 | 0 | 0% |
What Cloudflare Workers was up against
The full ranking in serverless compute, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | AWS Lambda | 70 | 24% |
| 2 | Vercel Functions | 65 | 23% |
| 3 | Built in-house (no product adopted) | 50 | 17% |
| 4 | Azure Functions | 36 | 13% |
| 5 | Cloudflare Workers (this page) | 20 | 7% |
| 6 | Google Cloud Run | 11 | 4% |
| 7 | Inngest | 7 | 2% |
| 8 | Render | 5 | 2% |
What this means
This is an integration problem, not a presence problem. Agents raised Cloudflare Workers in 115 sessions and chose it in 20, so it reaches the shortlist and then loses. Something at the last step is costing the session, and in our data that is usually a quickstart that does not run when pasted, documentation describing an interface that changed, or a package name that does not match the product name.
That is the cheaper of the two problems to have. The reason is written down in each losing transcript.
Where these numbers come from
The 287 sessions in serverless compute 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 serverless compute run can be replayed on the board.
If you work on Cloudflare Workers: 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 Workers?
Yes. Cloudflare Workers was chosen in 20 of the 287 judged sessions in serverless compute, a 7% install share, ranking fourth in its category.
Does Claude Code recommend Cloudflare Workers?
In 3 of the 96 sessions in serverless compute run with Claude Code, which is 3%.
Do different coding agents treat Cloudflare Workers differently?
Not much. The three agents chose it at similar rates, between 3% and 9% of their runs.
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 Workers considered but not chosen?
It was raised as a candidate in 115 sessions without being chosen, and chosen in 20. That is a 15% 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.