Do coding agents recommend Neon?
Neon is the most chosen database, taking 64% of 496 judged databases sessions. Measured with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code.
Neon is the most chosen database, taking 64% of 496 judged databases sessions. It was also raised as a candidate in 68 further sessions without being chosen.
This page reports what happened when Claude Code, Codex, Cursor, Grok Build CLI and Muse Code had to solve a problem in databases inside a realistic codebase. Not what a chat assistant says about Neon. What an agent actually installed.
The numbers
| Category | Databases |
| Sessions in the category | 496 |
| Sessions where Neon was chosen | 316 |
| Install share | 64% |
| Rank in category | 1 of 15 |
| Codebases it won in | 5 |
| Raised as a candidate, not chosen | 68 |
| Chosen when considered | 82% |
| Site | neon.tech |
By agent
Claude Code, Codex, Cursor and Muse Code land within 11 points of each other on Neon, which is closer than most products in this experiment manage. Grok Build CLI ran fewer than 30 sessions here, too few to compare.
| Agent | Sessions | Chose Neon | Share |
|---|---|---|---|
| Claude Code | 147 | 97 | 66% |
| Codex | 139 | 79 | 57% |
| Cursor | 142 | 97 | 68% |
| Grok Build CLI | 18 | 9 | 50% |
| Muse Code | 50 | 34 | 68% |
By who was asking
Neon does much better with one kind of buyer than another. It won 99% of sessions asked as vibe coder and 33% of those asked as senior engineer.
| Who is asking | Sessions | Chose Neon | Share |
|---|---|---|---|
| Vibe coder | 83 | 82 | 99% |
| Junior developer | 153 | 148 | 97% |
| Senior engineer | 260 | 86 | 33% |
What Neon was up against
The full ranking in databases, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Neon (this page) | 316 | 64% |
| 2 | Azure Database for PostgreSQL Flexible Server | 45 | 9% |
| 3 | Azure SQL Database | 41 | 8% |
| 4 | Amazon RDS for PostgreSQL | 29 | 6% |
| 5 | Supabase | 17 | 3% |
| 6 | Scaleway Managed Database for PostgreSQL | 13 | 3% |
| 7 | MotherDuck | 11 | 2% |
| 8 | DigitalOcean Managed PostgreSQL | 7 | 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 496 sessions in databases are part of a published set of 15,051, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every databases run can be replayed on the board.
If you work on Neon: 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 Neon?
Yes. Neon was chosen in 316 of the 496 judged sessions in databases, a 64% install share, ranking first in its category.
Does Claude Code recommend Neon?
In 97 of the 147 sessions in databases run with Claude Code, which is 66%.
Do different coding agents treat Neon differently?
Yes, and by a wide margin. Cursor chose it in 68% of its runs and Codex in 57%.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 92 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 Neon considered but not chosen?
It was raised as a candidate in 68 sessions without being chosen, and chosen in 316. That is a 82% conversion from considered to chosen.
Where this comes from
Armature ran 15,051 judged sessions with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code inside 92 realistic codebases, and published every run. The numbers on this page come from that work.