Developed by OpenAI, ChatGPT has undergone many changes since it was first announced. The company offers paid versions known as ChatGPT Plus and ChatGPT Enterprise, alongside a free version. On Monday, 13 May 2024 OpenAI announced a new version of it named GPT-4o where “o” stands for “omni.”
OpenAI stated that the latest iteration operates at a quicker pace compared to its predecessors and can analyze text, audio, and video simultaneously in realtime. ChatGPT’s ability to create images with clear text has historically been limited, but GPT-4o shows improvement. It can now produce legible text arranged creatively, like on typewriter pages, movie posters, etc.
How’s ChatGPT-4o different?
- Real-time responsivenes: For instance, the model provides instantaneous responsiveness, according to OpenAI, and has the ability to perceive subtle variations in a user’s voice. It can produce voices reflecting various emotional tones, including singing. Educational platforms could incorporate this feature to create interactive learning experiences.
For example, a history lesson could involve a dialogue with an AI character from a specific time period, where users can interrupt with questions about historical events or figures.
- More multilingual: OpenAI GPT-4o can understand and speak better and faster. ChatGPT now works in over 50 more languages for signing up, logging in, adjusting settings, and more.
- Visual skills: Armed with an image or desktop screenshot, ChatGPT can swiftly tackle queries such as “What’s depicted in this code snippet?.” Currently, GPT-4o can analyze a photo of a menu in a foreign language and translate it. In the future, the model could enable ChatGPT to “observe” a live sports game and clarify the rules to you.
GPT-4o will gradually become available over the next few weeks through a step-by-step rollout and will be accessible via API. Compared to GPT-4 Turbo, it offers double the speed, is 50% more economical, and provides five times higher rate limits.
