OpenAI has released a substantial improvement to its widely used ChatGPT platform. This version incorporates GPT-4o, a strong new AI model that adds "GPT-4-class chat" to the app.
OpenAI CEO Mira Murati and coworkers demonstrated their latest flagship model, a real-time vocal chatbot that mimics human speech.
GPT-4 o, with its faster GPT-4 intelligence, is expected to revolutionize collaboration, making interaction more natural and easier in the future.
GPT-4o, an omni model that integrates speech, text, and visual, is claimed to be two times quicker and more efficient than its predecessor.
OpenAI's CTO Mira Murati announced that the ChatGPT platform will be free for all users, with paying customers having up to five times the capacity limits.
OpenAI is introducing new features to GPT-4o, allowing users to search the web, speak to the chatbot, and save information. These features will be available to ChatGPT Plus subscribers in the coming weeks.
GPT-4 o, with its visual abilities, can analyze desktop screenshots and process videos and pictures uploaded via an iPhone app, with a Windows equivalent soon.
The OpenAI demo focused on the company's workers addressing questions to the vocal ChatGPT, who answered with jokes and human-like conversation.
Real-time interaction: Unlike earlier models, GPT-4o supports back-and-forth dialog without waiting for the model to conclude its statement.
Harmonised speech synthesis: GPT-4o can synthesize and harmonize diverse voices, resulting in a more natural dialogue experience.
Sophisticated conversations: GPT-4o provides "normal" conversational exchanges, translations, and more, all with the sophistication anticipated of GPT-4 technology.