View all AI news articles

OpenAI Just Changed the Game: Meet GPT-4o, The AI That Talks, Sees, and Listens Like You!

May 17, 2024

Watch the Demonstration

For those interested in seeing GPT-4o in action, watch a presentation on YouTube showcasing the capabilities of this new model. Watch the demo here.

  • Introduction of GPT-4o by OpenAI.
  • Key features and enhancements in GPT-4o.
  • The expansion of OpenAI's product ecosystem.
  • Pricing and accessibility details for different user tiers.

The Launch of GPT-4o

OpenAI has recently unveiled GPT-4o, their latest generative AI model, marking a significant evolution from the previous GPT-4 version. According to Mira Murati, CTO of OpenAI, this new model boasts improvements in speed and expanded capabilities in handling text, images, and now, audio. The rollout of GPT-4o will occur in phases across various OpenAI platforms, including the consumer-favorite, ChatGPT. Notably, Murati emphasized that this enhancement will not incur additional costs for users.

Multimodal Capabilities and API Access

Celebrating GPT-4 Omni's Grand Entrance

In a bold move to democratize AI, OpenAI announced that the GPT-4o model is "natively multimodal," capable of understanding and generating content across voice, text, and images. This feature significantly enhances the versatility of applications built on OpenAI’s platform. Additionally, developers keen on experimenting with this new capability will find the API access more economical—costing half as much as the previous iteration, yet delivering double the speed.

Comparison with Previous Models

Prior to the official announcement of GPT-4o, there was speculation about OpenAI's next big project, with guesses ranging from a new AI search engine to an advanced voice assistant. However, the focus remained on refining the capabilities of GPT-4, leading to the birth of GPT-4o. Murati highlighted that while the intelligence level mirrors that of GPT-4, enhancements in text, vision, and audio processing set GPT-4o apart.

Enhanced User Interaction

One of the most appealing features of GPT-4o is its real-time interaction capability. Users can now engage with ChatGPT in a more dynamic way, including the ability to interrupt the AI mid-response. This feature, combined with the model's ability to detect emotional nuances and respond in various emotive tones, aims to enrich the user experience. Furthermore, ChatGPT can now act as a real-time audio and video assistant, featuring a refined and elegant voice that adds a touch of sophistication to interactions.

Vision and Multilingual Improvements

Coding with GPT-4 Omni

GPT-4o extends its prowess to visual inputs as well. For instance, it can analyze images to answer queries about content displayed on a screen or identify objects within a photo. Furthermore, OpenAI claims improvements in the model’s performance across 50 different languages, enhancing its global usability.

New Applications and UI Enhancements

Murati also announced the launch of a desktop version of ChatGPT and a refreshed user interface, initially available on macOS. This new application allows users to interact with ChatGPT alongside other programs seamlessly, indicating a stride towards more integrated user experiences. The desktop app and the UI updates are expected to be available to a broader audience in the coming weeks, with a Windows version slated for release later in the year.

OpenAI’s GPT Store Goes Public

Omni Assistant: Real-Time Talks with GPT-4o

In addition to the software updates, OpenAI has made its GPT Store accessible to all users, free of charge. This platform enables users to create and share custom GPT models. Previously exclusive to paying customers, the store's opening to the public could significantly impact the custom AI application landscape.


OpenAI's GPT-4o introduces significant advancements in AI capabilities and user accessibility, setting a new benchmark for interaction between humans and AI. With these developments, OpenAI not only enhances the functionality of its existing products but also broadens the horizon for future AI applications.

Recent articles

View all articles