OpenAI reported to the GPT-4o, a model is multi-modal, which makes it a more natural interaction human-computer. It will then be available for everyone, including users who are aksesojnë ChatGPT free of charge. Without a subscription now is it possible to access the Store, GPT, and to use the GPT-be. The company is led by Sam Altman have also found application in the new macOS, and përditësoi interface to the internet to the chatbot.
GPT-4o, it says, see, and speak
All the news was made public by the Good Murad, Head of the Technology of OpenAI, in the event of the broadcast on YouTube. The most important thing is definitely the GPT-4o, where the letter o is the iniciali the omni all-in Latin). The latest version of the model for the generation of HE may be taking a combination of text, images and audio, as well as the introduction and production of text, images and audio, as well as the output.
OpenAI has published a series of videos to show the capabilities of the GPT-4o. The improvements compared to the previous year, especially when it comes to a conversation. The Voice Mode is currently in use by three models were statistically significant, a transcription of the audios in the text, it's the one to provide the answer in the text (p.moreover, it has. GPT-3.5, or GPT-4), and the conversion of text to audio. The processing requires up to 5,4 seconds delay.
For the GPT-4o, training a model, just that it can handle text, images, and audion, with a speed similar to a human being. He is, in fact, you can answer an incoming audio to a little more than 232 milliseconds, with an average of 320 milliseconds. Chatbot offers an interactive, very similar to that of the Samantës, asistentes a voice in the movie, the one with Joaquin Phoenix.
These new capabilities will be added in a Voice Mode to ChatGPT Plus in the coming weeks. GPT-4o, will be also be available soon on the user's ChatGPT Plus Team. The free ChatGPT will be updated with GPT-4o, yet with restrictions on the number of messages. Initially, you will have access to the introduction to the text and the image. How can we access the model via the API. Modes of entry (audio and video) will be available only to a group of limited partners.
GPT and GPT Store for ChatGPT for Free
In the event of a mbrëmshme, OpenAI announced several new features for the users of the paabonuar. You can also use the model of GPT-4o, and to receive answers from the web with a blog and Browse to the Bing to analyze the information in the documentation, and to create the graphics, the use of images, as well as access to upload your files, and take advantage of the functionality of the memory.
The user's use of the ChatGPT Free at last to enter the Store, GPT, and advantage of the GPT-be. However, the creation of a chatbot was the custom with the GPT Builder, remain exclusive to subscribers of the ChatGPT a Plus When it reached the threshold of the messages to the (unspecified), ChatGPT passes automatically to the model, GPT-3.5 to carry on a conversation.
The application for the macOS, and the interface of the new website
As a user of the cheapest as well as the customers will be able to interact with the ChatGPT through the application of the new macOS. With the shortcut Option + Space opens a window to a little above that of the other programs. Chatbot can provide answers to the questions that are related to what we “see” on the screen.
In addition to entering text, entering the vocal is available (through the Voice Mode by right clicking on the icon of a microphone in the bottom right. The capabilities of audio or video of the GPT-4o, will be added in the near future. The application is now available to users of the Plus version of the Windows will arrive later this year.
In the end, OpenAI showed the interface of the new ChatGPT on the web. The improvements are mainly concerned with the home screen and the appearance of the message.
Discussion about this post