Spread the love

OpenAI announces GPT-4o, real-time voice (and lots of new free features)

© OpenAI

While the GPT-5 model is still awaited, OpenAI presented some new features coming to ChatGPT, including many things previously reserved for ChatGPT Plus subscribers and which are becoming free. The first announcement concerns the firm's LLM star model, which upgrades to the GPT-4o version. Yes, there is some confusion, given that the last model was called GPT-4.5 Turbo.

However, what follows the 4 this time is a letter that means “Omni”, and that changes everything. Especially when you interact not with the keyboard… but by voice.

GPT-4o delivers real-time interactions for the first time

This is undoubtedly the most magical new feature of this new LLM model: through who knows what magic trick, OpenAI now manages to deliver instantaneous responses to prompts, even complex ones – or taking advantage, we will see, of the video.

This exploit allows us to completely rethink interactions with AI, and at the same time voice assistants. OpenAI has already been offering voice interactions for some time. But if you try with the GPT-4.5 Turbo model while it remains available, you will quickly see its limit.

There is indeed a fairly long delay between the moment you finish your question and the moment ChatGPT gives you an answer. Which can be very frustrating and makes products like the Humane AI Pin so imperfect. For pin, in particular, which depends on OpenAI, the arrival of the GPT-4o model looks like a game changer

GPT-4o is “omni”, which includes text, voice, photos, and live video

GPT-4o is an LLM covering the same dataset as the previous model. But the firm was able to better interweave its “senses” – namely text, voice, photo and video inputs. What this does: ChatGPT can now understand the world around it much better.

The most stunning thing is undoubtedly the arrival of what appears to be AI recognition on live video from your smartphone's sensors. Very concretely if you point at something and ask something related to what the app sees, ChatGPT can now respond promptly.

All in real real time which we will obviously have to test when all this is indeed available.

An application arrives on computer

OpenAI announces GPT-4o, real-time voice (and lots of new free features)

© OpenAI

OpenAI also did the demonstration of an application that arrives on a computer. We were able to see the Mac version which delivers, among other things, the ChatGPT voice function in real time – and a mode allowing ChatGPT to simply see your screen.

The speakers' demo focused on code, then on the interpretation of curves. The whole, quite impressive, should provide new possibilities for getting help from AI in the context of any activity – professional or not.

The app is available on Mac today. Windows version coming “later this year”.

Many things reserved for ChatGPT Plus are becoming free

D’more good news has been announced, especially for free users. The first is that the GPT-4 model is available to free users… plus connection to the web. The model allows, among other things, without paying ChatGPT Plus, to analyze data and create graphs.

It is also possible to submit photos to the model 4 and ask ChatGPT to analyze them. Uploading files to help you summarize, write or analyze them becomes possible at no cost. Free users can now access GPT and the GPT Store. As well as Memory, the function which allows you to use the history of your conversations.

The real-time voice function remains exclusive to paying users of GPT Plus and more expensive subscriptions.

The arrival of these new features will be gradual

ChatGPT-4o is rolling out now to ChatGPT Plus and Teams users – with an arrival noted in your account “within’a few weeks”. The new real-time Voice Mode is also coming in a few weeks “in alpha version”.

The new model is also arriving on the same schedule to free users but with lower limits. This remains evanescent for the moment: we simply know that paying users will have a limit of prompts up to 5 times higher than on free accounts. Team and Enterprise subscribers will benefit from “even higher” according to OpenAI.

  • OpenAI is currently deploying the GPT-4o model allowing real-time interactions.< /li>
  • The model also allows for amazing voice interactions, without the slightest delay.
  • Many things that were reserved for Plus subscribers become free.

📍 To not miss any news from Presse-citron, follow us on Google News and WhatsApp.

Teilor Stone

By Teilor Stone

Teilor Stone has been a reporter on the news desk since 2013. Before that she wrote about young adolescence and family dynamics for Styles and was the legal affairs correspondent for the Metro desk. Before joining Thesaxon , Teilor Stone worked as a staff writer at the Village Voice and a freelancer for Newsday, The Wall Street Journal, GQ and Mirabella. To get in touch, contact me through my teilor@nizhtimes.com 1-800-268-7116