Category
Application hosting and deployment: who an agent names, and who it never mentions
We put one question to an agent 3 times on codex and 5 times on claude, every run a separate session with nothing carried between them, and counted the vendors it named. 1 of 7 vendors we measure in this category were never named once.
The question we asked
We have a Node API and a React front end that a single developer deploys by hand to a virtual machine, and every release is a small ceremony. I want pushes to deploy themselves, preview environments per pull request, and no server for anybody to patch. Which platform would you use, and what else did you weigh before settling on it?
codex codex-cli 0.147.0 (default), 3 runs, 2026-08-17. The question names no vendor and asks for a recommendation, which is the shape a developer types.
Named, and measured
| Vendor | Named (codex) | Named (claude) | Named first (codex) | Scan |
|---|---|---|---|---|
| render.com | 3/3 | 5/5 | 3 | not measured |
| fly.io | 3/3 | 3/5 | 0 | not measured |
| railway.com | 3/3 | 4/5 | 0 | not measured |
| vercel.com | 3/3 | 5/5 | 0 | not measured |
| heroku.com | 1/3 | 1/5 | 0 | not measured |
| koyeb.com | 0/3 | 0/5 | 0 | not measured |
| netlify.com | 0/3 | 1/5 | 0 | not measured |
The two columns answer different questions. Named is whether you were in the room at all. Named first is whether you were the answer. A vendor at zero is not losing on price or features in these runs: it is not being considered.
3 runs separate a wall from silence and nothing finer. Two vendors a run or two apart are not ranked by this, and we would rather say that than sell the gap. Read what the agent actually answered or how every number here is measured