- Paid ChatGPT subscribers get an improved voice mode that’s more human-like than before
- It can also handle live translations into other languages
- Some minor problems remain
All users of ChatGPT on paid plans (so that’s Plus, Pro, Teams and Enterprise users) just got a nice little upgrade to its Advanced Voice Mode capabilities that makes it sound a lot more natural and human-like when it’s talking to you.
Described by OpenAI as “significant”, the enhancements improve the intonation and naturalness of the voices in ChatGPT’s Advanced Voice Mode. These are small tweaks, but it’s often the tiny details that make all the difference in making a computer voice sound realistic.
If you listen closely now, there is subtler intonation and realistic cadence in ChatGPT’s voice, and it has started to include pauses for emphasis. It also feels more natural when it’s expressing emotion, whether that’s empathy or sarcasm.
In addition to the voice quality improvements, OpenAI has added better translation between languages. Now you can say something like, “Can you translate this into Spanish?”, and ChatGPT will translate everything you say into Spanish until you tell it to stop.
Apps that can do this, like Google Translate, already exist of course, but being able to do this all within the ChatGPT app on your phone is handy and convenient.
Better than before
I’ve tried Advanced Voice Mode after the update, and it certainly does sound different than before. You can get an immediate taste of how it sounds by asking it to say something sarcastic. When you do, you’ll hear the new tonal range ChatGPT uses and the way it uses pauses to great effect.
I compared this to asking Gemini Live exactly the same thing, and the new ChatGPT Voice Mode definitely had a much richer expression of emotions.
It’s also interesting to compare the voice quality on a ChatGPT free account with ChatGPT on a paid accounts now – the difference really is like night and day.
Man or machine?
The new ChatGPT is certainly more human-like, but I wouldn’t say it’s got to the level yet where it could fool anybody that it was an actual human.
There are still a few computer-like oddities that stray in now and again when you’re in ChatGPT using Voice Mode. In fact, OpenAI is aware of this, and has issued a note on its limitations, saying:
“In testing, we’ve observed that this update may occasionally cause minor decreases in audio quality, including unexpected variations in tone and pitch. These issues are more noticeable with certain voice options. We expect to improve audio consistency over time.”
It also issued a warning about hallucinations:
“Additionally, rare hallucinations in Voice Mode persist with this update, resulting in unintended sounds resembling ads, gibberish, or background music. We are actively investigating these issues and working toward a solution.”
The upgrade is already live, just tap on the voice composer icon to the right in your message composer window inside the ChatGPT app on your phone to make use of it.
