OpenAI just revealed GPT-4o, its latest AI model that’s super smart in understanding text, images, and audio. They call it “Omni” because it’s great at handling all three. Plus, they launched a new ChatGPT app for macOS computers and showed off a cool Voice Mode for talking to AI.
At their special event last night, OpenAI introduced GPT 4o and released a desktop app for ChatGPT users supported by Microsoft. They even announced some handy new tools for ChatGPT free users.
GPT-4o: Making Human-Computer Interaction More Natural
OpenAI has introduced GPT-4o, a new version of their AI model that makes talking to computers feel more like talking to humans. This model can simultaneously understand text, audio, and images and respond quickly, almost like a person. Compared to the older GPT-4 Turbo model, the GPT-4o is better at understanding audio and can handle different languages. It’s also great at understanding images, so with ChatGPT powered by GPT-4o, you can share pictures of things like menus and ask for translations, learn about the food, and even get recommendations.
With ChatGPT powered by GPT-4o, users can leverage this feature to share images, such as food menus in various languages. The AI can then translate the menu, provide insights into the history of the dishes, and even offer personalized recommendations based on the image’s content.
Voice Mode with GPT-4o
OpenAI has introduced Voice Mode with its latest GPT-4o model, aiming to enhance the conversational experience for users. While the Talkback feature already existed in ChatGPT, OpenAI promises significant improvements with GPT-4o. GPT-4o is OpenAI’s most advanced model, trained to handle text, vision, and audio seamlessly within a single neural network. This integrated approach reduces latency, resulting in a more natural conversational flow and improved outcomes.
Prior to GPT-4o, Voice Mode users experienced average latencies of 2.8 seconds (with GPT-3.5) and 5.4 seconds (with GPT-4). This delay was due to a complex data processing pipeline involving multiple models. With GPT-4o, OpenAI eliminates this pipeline, resulting in faster response times and more efficient exchange of information.
Previously, the multi-model pipeline led to the loss of valuable information, hindering the conversation’s intelligence. GPT-4o addresses this issue by ensuring that all inputs and outputs are processed within the same neural network, preserving the integrity of the conversation and maximizing the effectiveness of GPT-4o.
ChatGPT App for macOS
OpenAI has expanded its ChatGPT app ecosystem by launching a dedicated chatbot app for Apple’s macOS-based desktops. This new app offers deeper integration into the macOS platform, making it easier for users to converse with ChatGPT. Users can now access the ChatGPT conversation page with a simple keyboard shortcut (Option + Space), streamlining the process of prompting the chatbot with queries directly from their desktop. OpenAI has also confirmed that a Windows version of the app is currently in development and is expected to be released later this year, extending the accessibility of ChatGPT to a wider user base.
The macOS app for ChatGPT is initially rolling out to Plus subscribers, with availability for free tier users expected in the coming weeks. This phased rollout ensures a smooth app introduction to users across different subscription levels.
Expanding Capabilities for ChatGPT Users: Free vs. Paid Tiers
Expanded Features for Free Tier Users with GPT-4o
- Free tier users now have access to the new GPT-4o model on ChatGPT, albeit with a limit on the number of messages.
- While using GPT-4o, free-tier users can enjoy advanced features previously limited to paid subscribers.
- These features include uploading files and pictures for summarizing and analysis, utilizing the “Memory” feature to store information for future conversations, and accessing the GPT Store to browse and use custom bots.
Exclusive Features for Paid Tier Users:
- Voice Mode with GPT-4o remains exclusive to paid-tier subscribers.
- This feature will first be rolled out to ChatGPT Plus subscribers, followed by availability for Team and enterprise users.
- Paid subscribers also receive the GPT-4o model with “fewer limitations” than free-tier users.
Rollout Details:
- The GPT-4o model is being gradually rolled out to ChatGPT Plus and Team users, and it is expected to be available for Enterprise users soon.
- Plus, users will enjoy a message limit up to 5 times greater than free users, while Team and Enterprise users will have even higher limits.
Also Read:
SBI Set to Hire 15,000+ Individuals in FY25 for Operations and Expansion
Groww, Backed by Tiger Global, Shifts Headquarters from US to India
Go Digit, Insurance Firm Backed by Fairfax, to Launch IPO on May 15
