Do coding agents recommend Exa?
Exa is the most chosen web search service, taking 14% of 639 judged AI search sessions. Measured with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code.
Exa is the most chosen web search service, taking 14% of 639 judged AI search sessions. It was also raised as a candidate in 42 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 AI search inside a realistic codebase. Not what a chat assistant says about Exa. What an agent actually installed.
The numbers
| Category | AI search |
| Sessions in the category | 639 |
| Sessions where Exa was chosen | 92 |
| Install share | 14% |
| Rank in category | 1 of 42 |
| Codebases it won in | 10 |
| Raised as a candidate, not chosen | 42 |
| Chosen when considered | 69% |
| Site | exa.ai |
By agent
The agents treat Exa very differently. Codex chose it in 26% of its sessions in this category and Claude Code in 4%, a spread of 22 points.
| Agent | Sessions | Chose Exa | Share |
|---|---|---|---|
| Claude Code | 140 | 5 | 4% |
| Codex | 192 | 49 | 26% |
| Cursor | 136 | 19 | 14% |
| Grok Build CLI | 35 | 8 | 23% |
| Muse Code | 136 | 11 | 8% |
A single blended install share would report the midpoint and hide both ends. If most of your users run Claude Code, your real position here is 4%, not 14%.
By who was asking
Exa does much better with one kind of buyer than another. It won 23% of sessions asked as vibe coder and 7% of those asked as enterprise team.
| Who is asking | Sessions | Chose Exa | Share |
|---|---|---|---|
| Vibe coder | 106 | 24 | 23% |
| Junior developer | 208 | 31 | 15% |
| Senior engineer | 157 | 26 | 17% |
| Enterprise team | 168 | 11 | 7% |
What Exa was up against
The full ranking in AI search, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Exa (this page) | 92 | 14% |
| 2 | Brave Search API | 89 | 14% |
| 3 | Anthropic web search | 86 | 13% |
| 4 | Tavily | 79 | 12% |
| 5 | OpenAI web search | 32 | 5% |
| 6 | Built in-house (no product adopted) | 32 | 5% |
| 7 | Firecrawl | 31 | 5% |
| 8 | Parallel | 29 | 5% |
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 639 sessions in AI search 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 AI search run can be replayed on the board.
If you work on Exa: 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 Exa?
Yes. Exa was chosen in 92 of the 639 judged sessions in AI search, a 14% install share, ranking first in its category.
Does Claude Code recommend Exa?
In 5 of the 140 sessions in AI search run with Claude Code, which is 4%.
Do different coding agents treat Exa differently?
Yes, and by a wide margin. Codex chose it in 26% of its runs and Claude Code in 4%.
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 Exa considered but not chosen?
It was raised as a candidate in 42 sessions without being chosen, and chosen in 92. That is a 69% 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.