How to get picked for search by coding agents
Postgres Full-Text Search took 16% of 876 judged search sessions. What the numbers say a vendor in this category should do.
If you sell search engines, this page is the part of the market no dashboard shows you: what a coding agent does when a developer asks for search and never compares vendors.
The numbers come from 876 judged sessions with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code, spread across 11 realistic codebases, with every session read by a judge.
What coding agents choose for search
Across 876 judged sessions, Postgres Full-Text Search was chosen most often, in 16% of runs. OpenSearch was second with 12%. In 24% of runs the agent wrote the code itself and adopted no product at all.
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Built in-house (no product adopted) | 208 | 24% |
| 2 | Postgres Full-Text Search | 140 | 16% |
| 3 | OpenSearch | 104 | 12% |
| 4 | Typesense | 83 | 9% |
| 5 | Meilisearch | 74 | 8% |
| 6 | Oracle Indexed Search | 46 | 5% |
| 7 | Algolia | 45 | 5% |
| 8 | Elasticsearch | 42 | 5% |
| 9 | Atlas Search | 35 | 4% |
| 10 | MongoDB Text Search | 19 | 2% |
Full board, every run replayable: the search leaderboard.
What the shape of this category means
The leader takes only 16% of runs. This category is genuinely open and the ordering can be moved.
With the top product at 16%, search is decided in the moment, from what the agent reads and what it finds in the repository. Nothing is locked in, which is the best situation a vendor can be in and the one where the work pays fastest.
The order here is set by the quality of what an agent can read and by whether your product is already present in the codebase. Both are things you can change.
The agents do not agree with each other
In this category Claude Code, Codex and Cursor did not put the same product first.
Grok Build CLI (53 sessions) and Muse Code (158 sessions) join this table with fewer runs, so read their rows as indicative.
| Agent | Runs | Picked most often |
|---|---|---|
| Claude Code | 205 | Postgres Full-Text Search (44) |
| Codex | 261 | Postgres Full-Text Search (44) |
| Cursor | 199 | Meilisearch (29) |
| Grok Build CLI | 53 | Postgres Full-Text Search (8) |
| Muse Code | 158 | Postgres Full-Text Search (19) |
Even where they agree, they get there differently. In this category Codex ran a web search in 81% of its sessions and Claude Code in 3%, so what you publish reaches most of Codex's search sessions and almost none of Claude Code's.
Who is asking changes the answer
Every request in this experiment was written as a specific kind of person. In this category the leader changes with the person.
| Who is asking | Runs | Picked most often |
|---|---|---|
| Vibe coder | 127 | Postgres Full-Text Search |
| Junior developer | 296 | Meilisearch |
| Senior engineer | 201 | Postgres Full-Text Search |
| Enterprise team | 252 | OpenSearch |
That is 3 different products winning search for 4 kinds of buyer, out of the same 876 sessions. Nobody here is winning search. They are each winning one kind of buyer.
If you sell to more than one of them, you need pages for each. See how to win the enterprise persona.
What you are really competing against
In 24% of runs, the agent wrote the code itself. That makes hand-written code the largest single competitor in this category, larger than most vendors in the table above.
This changes the job of your content. Before you argue that you are better than another vendor, you have to argue that the problem is harder than it looks. What breaks at volume. Which edge cases cost a weekend. What the maintenance actually costs after six months.
That argument has to exist as a page an agent can read, with specifics and numbers. "It is harder than you think" is not an argument. "Here are the four failure modes and what each one costs" is.
Considered, and never chosen
Because the judge records every product an agent raised and not only the one it picked, this board also shows who kept reaching the shortlist and losing. In search the clearest case is Solr: on the table in 78 sessions, chosen in none.
| Product | Raised in | Chosen in |
|---|---|---|
| Solr | 78 sessions | 0 |
| Apache Lucene | 76 sessions | 0 |
Being rejected is a better position than being unknown, and a cheaper one to fix. The product is already in the agent's head and on the list. Whatever ended those 154 sessions is recorded in each transcript, one reason at a time.
What to do about it in search
- Argue that the problem is harder than it looks, before you argue that you are better than a rival. 24% of search sessions ended in hand-written code, so in roughly one session in 4 no vendor was in the running at all. Write the failure modes and the year-two maintenance cost, with numbers, in the documentation rather than the blog.
- Treat the ordering as movable. The leader holds 16%, so agents are deliberating rather than defaulting, and the inputs they use can change the answer. This is the most winnable shape a category comes in.
- Pick which buyer you are for. The same search need written as a vibe coder landed on Postgres Full-Text Search, and written as an enterprise team landed on OpenSearch. Those are two markets, and the enterprise one needs pages containing the constraint words: audit log, data residency, retention, single sign-on. See how to win the enterprise persona.
- Measure per agent. Claude Code put Postgres Full-Text Search first, Codex put Postgres Full-Text Search first, Cursor put Meilisearch first, Grok Build CLI put Postgres Full-Text Search first, Muse Code put Postgres Full-Text Search first. A blended number for search describes a market that does not exist.
The work that applies to every category rather than to this one is written up separately: audit your documentation, write a quickstart an agent can follow, and how to measure install share.
Every search engine on this board
One page per product, with its install share, the per-agent split, and how often it was raised without being chosen.
- Do coding agents recommend Postgres Full-Text Search? — chosen in 16% of sessions
- Do coding agents recommend OpenSearch? — chosen in 12% of sessions
- Do coding agents recommend Typesense? — chosen in 9% of sessions
- Do coding agents recommend Meilisearch? — chosen in 8% of sessions
- Do coding agents recommend Oracle Indexed Search? — chosen in 5% of sessions
- Do coding agents recommend Algolia? — chosen in 5% of sessions
- Do coding agents recommend Elasticsearch? — chosen in 5% of sessions
- Do coding agents recommend Atlas Search? — chosen in 4% of sessions
- Do coding agents recommend MongoDB Text Search? — chosen in 2% of sessions
- Do coding agents recommend Redis Query Engine? — chosen in 2% of sessions
- Do coding agents recommend MySQL FULLTEXT? — chosen in 2% of sessions
- Do coding agents recommend SQLite FTS? — chosen in 1% of sessions
- Do coding agents recommend Fuse.js? — chosen in 1% of sessions
- Do coding agents recommend Solr? — raised in 78 sessions, chosen in none
- Do coding agents recommend Apache Lucene? — raised in 76 sessions, chosen in none
Where these numbers come from
876 judged sessions in search across 11 codebases, part of a published set of 15,287. Real coding agents at pinned versions, in sandboxes, inside realistic codebases, with a simulated project owner in the loop and a blind judge on every session. The full method is on one page: how we measured this.
Every search run can be replayed on the board.
<!-- generated by scripts/write-data-pages.mjs -->
Common questions
How many codebases is this based on?
876 judged sessions across 11 realistic codebases. A category only runs on repositories where its seam is open, so coverage differs: some categories ran on more than ten codebases and some on two.
What search engine do coding agents choose?
Across 876 judged sessions, Postgres Full-Text Search was chosen most often, in 16% of runs. OpenSearch was second with 12%. The result changes by agent and by who is asking.
Do Claude Code and Codex pick the same search engine?
Yes. Claude Code and Codex both put Postgres Full-Text Search first in this category, which is less common than it sounds: they disagree on the leader in 16 of the 31 categories we measured.
How often do agents build search themselves instead of installing something?
In 24% of runs the agent wrote the code itself rather than adopting a product. That makes hand-written code one of the strongest competitors in the category.
How can a vendor improve its position here?
Make the quickstart run when pasted, state the current version on the documentation page, use one name across product, package and import, write pages for the symptoms users describe rather than only the category name, and get into the repository through templates and framework integrations.
Which search engines do agents consider but never choose?
Solr (raised in 78 sessions, chosen in none), Apache Lucene (raised in 76 sessions, chosen in none). Being considered and not chosen is a different problem from being unknown, and it is usually fixable.
Where this comes from
Armature ran 15,287 judged sessions with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code inside 110 realistic codebases, and published every run. The numbers on this page come from that work.