codex - 💡(How to fix) Fix Japanese voice input frequently appends unexpected trailing punctuation sequence [1 participants]

What version of the Codex App are you using (From “About Codex” dialog)?

26.429.61741 (2429)

What subscription do you have?

ChatGPT Pro

What platform is your computer?

macOS Darwin 25.4.0 arm64 arm

What issue are you seeing?

When I use voice input in the Codex app while speaking Japanese, the transcription frequently appends an unexpected trailing character sequence at the end of the dictated text.

The unexpected trailing sequence appears to be the following Unicode code points:

• U+3011 RIGHT WHITE SQUARE BRACKET
• U+3002 IDEOGRAPHIC FULL STOP
• U+FF1B FULLWIDTH SEMICOLON

This has happened dozens of times during normal use, so it appears to be reliably reproducible rather than a one-off transcription mistake.

I am speaking Japanese naturally, and the rest of the transcription is usually correct. The issue is specifically that this extra punctuation/bracket sequence is appended at the end of the voice input.

I also noticed that when I tried to mention the exact problematic character sequence in a Codex conversation, the assistant response appeared to stop or truncate around that sequence. This may indicate a separate rendering/input handling issue related to the same characters, but the primary issue is the voice transcription adding them unexpectedly.

What steps can reproduce the bug?

1. Open the Codex app on macOS.
2. Start a conversation/thread.
3. Use voice input/dictation.
4. Speak in Japanese.
5. Submit the transcribed prompt.
6. Observe that an unexpected trailing sequence is appended to the end of the Japanese transcription.

This has occurred dozens of times for me during normal Japanese voice input.

What is the expected behavior?

No response

Additional information

No response