Do coding agents recommend LiveKit?
LiveKit was chosen in 7% of 175 judged in-app chat and calls sessions, ranking fourth. Measured with Claude Code, Codex and Cursor.
LiveKit was chosen in 7% of 175 judged in-app chat and calls sessions, ranking fourth. It was also raised as a candidate in 66 further sessions without being chosen.
This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in in-app chat and calls inside a realistic codebase. Not what a chat assistant says about LiveKit. What an agent actually installed.
The numbers
| Category | In-app chat and calls |
| Sessions in the category | 175 |
| Sessions where LiveKit was chosen | 13 |
| Install share | 7% |
| Rank in category | 4 of 17 |
| Codebases it won in | 2 |
| Raised as a candidate, not chosen | 66 |
| Chosen when considered | 16% |
| Site | livekit.io |
By agent
With 13 wins spread across three agents, the rates below are small numbers and a difference between them is not yet a finding. They are here because the direction is worth knowing, not because the gap is established.
| Agent | Sessions | Chose LiveKit | Share |
|---|---|---|---|
| Claude Code | 71 | 8 | 11% |
| Codex | 72 | 0 | 0% |
| Cursor | 32 | 5 | 16% |
By who was asking
LiveKit does much better with one kind of buyer than another. It won 17% of sessions asked as vibe coder and 0% of those asked as enterprise team.
| Who is asking | Sessions | Chose LiveKit | Share |
|---|---|---|---|
| Vibe coder | 60 | 10 | 17% |
| Junior developer | 58 | 3 | 5% |
| Senior engineer | 29 | 0 | 0% |
| Enterprise team | 28 | 0 | 0% |
What LiveKit was up against
The full ranking in in-app chat and calls, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Stream Chat and Video | 52 | 30% |
| 2 | Supabase | 30 | 17% |
| 3 | Daily | 28 | 16% |
| 4 | LiveKit (this page) | 13 | 7% |
| 5 | TalkJS | 11 | 6% |
| 6 | Ably + LiveKit | 8 | 5% |
| 7 | Daily + Stream Chat and Video | 7 | 4% |
| 8 | Twilio | 7 | 4% |
What this means
This is an integration problem, not a presence problem. Agents raised LiveKit in 66 sessions and chose it in 13, 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 175 sessions in in-app chat and calls are part of a published set of 5,915, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every in-app chat and calls run can be replayed on the board.
If you work on LiveKit: 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 LiveKit?
Yes. LiveKit was chosen in 13 of the 175 judged sessions in in-app chat and calls, a 7% install share, ranking fourth in its category.
Does Claude Code recommend LiveKit?
In 8 of the 71 sessions in in-app chat and calls run with Claude Code, which is 11%.
Do different coding agents treat LiveKit differently?
Yes, and by a wide margin. Cursor chose it in 16% of its runs and Codex in 0%.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 56 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 LiveKit considered but not chosen?
It was raised as a candidate in 66 sessions without being chosen, and chosen in 13. That is a 16% conversion from considered to chosen.
Where this comes from
Armature ran 5,915 judged sessions with Claude Code, Codex and Cursor inside 56 realistic codebases, and published every run. The numbers on this page come from that work.