Voice Chat speaks Chinese responses too fast on macOS app
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 125k
- Forks
- 19.4k
- PR merge metrics
- PR metrics pending
Description
What version of the Codex App are you using (From “About Codex” dialog)?
26.818.41509
What subscription do you have?
Pro 5x
What platform is your computer?
Darwin 25.5.0 arm64 arm
What issue are you seeing?
When using Voice Chat in the Codex macOS app, Chinese spoken responses are played/spoken much faster than expected. The text response itself is Chinese and correct, but the audio pace is unnaturally fast and hard to understand.
Environment
App: Codex desktop app
Language used: Chinese
Feature: Voice Chat
Task: Ask Codex to summarize the current screen
Actual Behavior
Codex responds in Chinese, but the spoken audio is very fast. The pace feels unnatural and makes the response difficult to understand.
What steps can reproduce the bug?
-
Open the Codex macOS app.
-
Start Voice Chat.
-
Speak to Codex in Chinese.
-
Ask it to summarize the current screen content.
-
Listen to the Chinese audio response.
What is the expected behavior?
Chinese Voice Chat responses should be spoken at a natural, intelligible pace, similar to normal conversational Chinese.
Additional information
An audio recording demonstrating this issue
Note
- When using Voice Chat in ChatGPT on iPad, Chinese spoken responses sound normal and are easy to understand. The issue appears specific to the Codex macOS app experience.
- This may be an upstream Realtime voice pacing issue or a missing Voice Chat speech-pace control, but it is surfaced through the Codex macOS app
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the issue in the Codex macOS app using the listed version, Darwin environment, Chinese input, and Voice Chat screen-summary flow. Compare the spoken pacing with the attached recording and determine whether the behavior is local to the app or an upstream Realtime voice issue. Done means Chinese responses play at a natural, intelligible pace; no source file or test is named.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- macos, rust
- Domain
- audio-video-rtc, desktop
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Needs clarification
- Newbie friendliness
- 38/100