openai / openai/codex

Support Voice mode in Codex Desktop with Azure OpenAI as the model provider

Open
#44,131 5 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

app azure custom-model enhancement
Dominant language
Rust
Stars
125k
Forks
19.4k
PR merge metrics
PR metrics pending

Description

What variant of Codex are you using?

Codex Desktop app (macOS)

What feature would you like to see?

Please support full Voice mode when a Codex task uses Azure OpenAI as its model provider.

I want to speak with Codex, hear its responses, and work on the same task while keeping my Azure-backed model selected. Switching the task to an OpenAI/ChatGPT-backed model should not be necessary just to use Voice mode.

The requested experience is a two-way voice conversation, beyond speech-to-text dictation in the composer.

Expected behavior
  • Start and continue a voice conversation on a task configured with an Azure OpenAI provider.
  • Keep the task's selected Azure model, context, and tools throughout the conversation.
  • Provide a documented setup for voice authentication and routing. If Voice uses a separate service or ChatGPT sign-in, make that dependency explicit and allow it to coexist with the Azure task provider.
  • Clearly explain any unsupported configuration or missing prerequisite in the UI.

It would also be useful to support an Azure-hosted audio/realtime deployment where compatible, with separate voice-provider settings if needed. The primary request is a supported way to use Voice mode while the coding task remains on Azure.

Additional information

Related reports:

  • #28149 describes a Codex Desktop dictation authentication failure with an Azure provider.
  • #13097 requests custom-provider support for CLI voice transcription.

These cover related transcription paths. This request focuses on full conversational Voice mode in the Desktop app with an Azure-backed task.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

The issue names no implementation files, tests, or entry points. Start by locating the Codex Desktop Voice mode and Azure provider integration, then read related issues #28149 and #13097 for authentication and transcription constraints. Done means a two-way voice conversation works with an Azure-backed task, preserves its model, context, and tools, and documents or clearly reports provider prerequisites.

Written by the indexing model from the issue text.

Assessment

Tech stack
azure, macos, rust
Domain
audio-video-rtc, cloud, desktop
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.