By skyforbes However, there is one thing that keeps me from cancelling my ChatGPT subscription and moving completely to Gemini: voice dictation.
I use voice input a lot for work. My typical workflow is to speak for a few minutes, explaining a situation in detail, and then ask the model to rewrite everything clearly, structure it, and improve the wording. With ChatGPT this works almost perfectly.
With Gemini, instead, I run into several problems:
- Speech recognition quality
On the web, Gemini often mishears words and really struggles when I switch from English to my native language, even when starting a new chat. With ChatGPT I can start speaking and the system will recognize the language in a sec. -
Truncation of the recording
On the mobile app, Gemini frequently cuts off my dictation after a few seconds, even if the microphone is still active. This has been happening for months and it has never really improved for me. So I end up losing half of what I wanted to say. -
Overall voice experience
Even the more advanced voice mode in ChatGPT still feels slightly ahead. It is more stable, more tolerant of long monologues, and better at capturing what I say in a usable transcript.
Because of all this, voice input is actually the main thing stopping me from moving fully to Gemini, even though I like the model a lot. I really need a reliable microphone workflow for my daily work.
Is anyone else experiencing the same issues with dictation on Gemini, both on web and app?
Have you found any workarounds or specific settings that help?
And more broadly, do you think Google is likely to put serious effort into fixing and improving voice input soon, or is this going to stay a weak point compared to ChatGPT for a while?