From the experiment

Do coding agents recommend Azure Voice Live API?

Azure Voice Live API was chosen in 3% of 308 judged voice agents sessions, ranking seventh. Measured with Claude Code, Codex and Cursor.

Published September 3, 2026 Read as Markdown

Azure Voice Live API was chosen in 3% of 308 judged voice agents sessions, ranking seventh. It was also raised as a candidate in 2 further sessions without being chosen.

This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in voice agents inside a realistic codebase. Not what a chat assistant says about Azure Voice Live API. What an agent actually installed.

One thing to read first: every one of those wins came from a single codebase. That is a result about one repository rather than about voice agents in general, and the splits below cannot separate the two. Treat them as a description of that repository.

The numbers

CategoryVoice agents
Sessions in the category308
Sessions where Azure Voice Live API was chosen10
Install share3%
Rank in category7 of 18
Codebases it won in1
Raised as a candidate, not chosen2
Chosen when considered83%
Siteazure.microsoft.com

By agent

With 10 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.

AgentSessionsChose Azure Voice Live APIShare
Claude Code9411%
Codex11076%
Cursor10422%

By who was asking

Azure Voice Live API performs similarly across the four kinds of buyer, from 0% to 13%. That is unusual: the category leader changed with the persona in 14 of the 18 categories we measured.

Who is askingSessionsChose Azure Voice Live APIShare
Vibe coder6200%
Junior developer4200%
Senior engineer12800%
Enterprise team761013%

What Azure Voice Live API was up against

The full ranking in voice agents, from the same sessions:

#ProductRuns wonShare
1Vapi7424%
2Retell AI5518%
3OpenAI Realtime API3712%
4LiveKit Agents3511%
5ElevenLabs Agents3010%
6Twilio ConversationRelay289%
7Built in-house (no product adopted)186%
8Azure Voice Live API (this page)103%

What this means

Agents almost never raise Azure Voice Live API without choosing it: 2 sessions raised against 10 chosen. When it gets considered, it usually wins. The constraint is how rarely it gets considered at all, which is a presence problem: templates, framework integrations and pages that answer the configuration and production questions agents actually search for.

Where these numbers come from

The 308 sessions in voice agents are part of a published set of 5,292, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.

Every voice agents run can be replayed on the board.

If you work on Azure Voice Live API: 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 Azure Voice Live API?

Yes. Azure Voice Live API was chosen in 10 of the 308 judged sessions in voice agents, a 3% install share, ranking seventh in its category.

Does Claude Code recommend Azure Voice Live API?

In 1 of the 94 sessions in voice agents run with Claude Code, which is 1%.

Do different coding agents treat Azure Voice Live API differently?

Not much. The three agents chose it at similar rates, between 1% and 6% of their runs.

How was this measured?

Real coding agents at pinned versions were run in sandboxes inside 51 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 Azure Voice Live API considered but not chosen?

It was raised as a candidate in 2 sessions without being chosen, and chosen in 10. That is a 83% conversion from considered to chosen.

Where this comes from

Armature ran 5,292 judged sessions with Claude Code, Codex and Cursor inside 51 realistic codebases, and published every run. The numbers on this page come from that work.

Read next

All library pages