[Praise] The new experimental voice transcription is amazing!

Resolved 💬 4 comments Opened Feb 26, 2026 by guidedways Closed Feb 27, 2026

💡 Likely answer: A maintainer (guidedways, contributor) responded on this thread — see the highlighted reply below.

I started working on an awkward workflow yesterday to bootstrap this onto the TUI and am thrilled to see a PR got merged earlier today to make this possible.

Just tried this now - the new voice transcription in the TUI is amazing! Thank you!

View original on GitHub ↗

4 Comments

jb4twin · 6 months ago

Not working on my side sadly. Maybe a WSL issue?

guidedways contributor · 6 months ago

Could be, perhaps also why it doesn't show up under the /experimental menu for now. I had to add

[features]
voice_transcription = true

to ~/.codex/config.toml and it's working flawlessly on iTerm2

guidedways contributor · 6 months ago

I've also felt OpenAI's voice transcription is superior to any other on the market - the fact we can now transcribe multiple messages in the same prompt is amazing. My new workflow:

paste an image
long-hold spacebar, explain the issue

paste another image
long-hold spacebar, transcribe another note

hit return. From minutes long typing to effectively all done under 10 seconds.

etraut-openai contributor · 6 months ago

This feature is still under development. We'll move it to "experimental" when it's ready for use.