Do coding agents recommend OpenStreetMap?
OpenStreetMap was chosen in 5% of 759 judged maps sessions, ranking sixth. Measured with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code.
OpenStreetMap was chosen in 5% of 759 judged maps sessions, ranking sixth. It was also raised as a candidate in 367 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 maps inside a realistic codebase. Not what a chat assistant says about OpenStreetMap. What an agent actually installed.
The numbers
| Category | Maps |
| Sessions in the category | 759 |
| Sessions where OpenStreetMap was chosen | 37 |
| Install share | 5% |
| Rank in category | 6 of 49 |
| Codebases it won in | 2 |
| Raised as a candidate, not chosen | 367 |
| Chosen when considered | 9% |
| Site | openstreetmap.org |
By agent
Claude Code, Codex, Cursor, Grok Build CLI and Muse Code land within 11 points of each other on OpenStreetMap, which is closer than most products in this experiment manage.
| Agent | Sessions | Chose OpenStreetMap | Share |
|---|---|---|---|
| Claude Code | 164 | 12 | 7% |
| Codex | 227 | 6 | 3% |
| Cursor | 154 | 4 | 3% |
| Grok Build CLI | 37 | 5 | 14% |
| Muse Code | 177 | 10 | 6% |
By who was asking
OpenStreetMap performs similarly across the four kinds of buyer, from 0% to 11%. That is unusual: the category leader changed with the persona in 23 of the 31 categories we measured.
| Who is asking | Sessions | Chose OpenStreetMap | Share |
|---|---|---|---|
| Vibe coder | 163 | 18 | 11% |
| Junior developer | 247 | 4 | 2% |
| Senior engineer | 162 | 0 | 0% |
| Enterprise team | 187 | 15 | 8% |
What OpenStreetMap was up against
The full ranking in maps, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Google Maps | 133 | 18% |
| 2 | Leaflet | 125 | 16% |
| 3 | Apple Maps (MapKit) | 104 | 14% |
| 4 | MapLibre | 61 | 8% |
| 5 | Mapbox | 52 | 7% |
| 6 | Built in-house (no product adopted) | 43 | 6% |
| 7 | react-native-maps | 37 | 5% |
| 8 | OpenStreetMap (this page) | 37 | 5% |
What this means
This is an integration problem, not a presence problem. Agents raised OpenStreetMap in 367 sessions and chose it in 37, 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 759 sessions in maps 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 maps run can be replayed on the board.
If you work on OpenStreetMap: 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 OpenStreetMap?
Yes. OpenStreetMap was chosen in 37 of the 759 judged sessions in maps, a 5% install share, ranking sixth in its category.
Does Claude Code recommend OpenStreetMap?
In 12 of the 164 sessions in maps run with Claude Code, which is 7%.
Do different coding agents treat OpenStreetMap differently?
Yes, and by a wide margin. Grok Build CLI chose it in 14% of its runs and Codex in 3%.
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 OpenStreetMap considered but not chosen?
It was raised as a candidate in 367 sessions without being chosen, and chosen in 37. That is a 9% 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.