I have been a GPT-Plus user since October I believe and I just got the news yesterday of the release of the new version omni. This new AI Model is able to take multiple input like text, image, voice and reply with similar output.
OpenAI has reduce to latency by a some milliseconds but it really make the difference here because it is becoming close to a normal human response time so you can now have a vocal conversation and that is impressive.
You have multiple choices of voices
The pace at which it is getting new model out is alarming, for me at least I can see a new paradigm shift happening again like when the first public version went out. We barely had time to accomadte to it and see how that could help use and OpenAI juste release an even more impressive version that could make GPT-4o the an well performing virtual assistant.
I am thrilled about finding out the new possibility and how I can leverage that for organising my life and help me in my daily tasks.
In term of UI it didn't really change, you can now select the new model and I know that with a subscription you can get 80 message per 3 hours compare to less then 20 with the free model which is already nice if you don't want to pay because you still have access to the latest version which wasn't the case when GPT-4 got out.
The experience real way more immersive now even with the ability to have real time conversation, I still do encounter some error when the stream isn't responding with the answer so it can feel slow but actually is is just stuck and it is best to refresh your browser and/or open a new chat.
I find that the hardest part is actually knowing what to test because you don't really know the capabilities and I will tend to use it like I do instead of trying new thing, that is why watching the open AI video is a plus that where I discovered that you could give it access to the camera so it can solve math problem on a piece of paper and explain to you like a teacher.
You can checkout the video on YouTube or just try it yourself first
