OpenAI Rolls Out GPT-4o Model with Faster Speeds and Lower Pricing

OpenAI has begun rolling out its newest version of ChatGPT, GPT-4o, with faster speeds and lower pricing.

OpenAI has begun rolling out its latest generative AI model, GPT-4o. Announced on May 13, the update is expected to improve upon ChatGPT’s capabilities in formats beyond just text. The “o” in GPT-4o stands for “omni,” the company said. This refers to GPT-4o’s ability to understand and process not only text but also speech and video.

During the streamed presentation, OpenAI CTO Mira Murati said the increased capabilities in voice and vision are “incredibly important, because we’re looking at the future of interaction between ourselves and machines.”

Improving ChatGPT’s Capabilities as a Voice Assistant

OpenAI says GPT-4o is supposed to make the voice mode in the ChatGPT app more like an assistant than ever before, allowing users to provide real-time responsiveness. Using a new text-to-speech model, the mobile app should allow you to have a back-and-forth conversation that even allows you to interrupt the chatbot to correct it or clarify what you want.

According to OpenAI, you can use ChatGPT to do things like request a bedtime story, settle a dinner table debate, or just have a conversation. Our own testing hasn’t proven so natural, unfortunately. In fact, I’ve yet to get the app on my phone to do anything in Voice Mode. It only reports back that its servers are overloaded and to try again later. Major buzzkill.

Photo and Video Analysis That’s Still Mediocre At Best

The new model is also supposed to allow for better analysis of images and videos. Here, too, it clearly needs further refinement and fixing.

OpenAI Video Analysis Hallucination

I tried uploading a video of my dog playing in the living room. When I finally coaxed an analysis of the video, it erroneously said the video was outdoors, with trees, grass, and other foliage. Ummm, no.

Wider Availability of the Latest Model Than Ever Before

The good news, maybe, is that GPT-4o is available across all tiers, even for free users. In the past, OpenAI’s free tier only had access to the previous version of the model. Users on the Plus tier get up to a five-fold higher message limit, and API pricing for 4o is 50% of the pricing for GPT-4 Turbo.

GPT-4o has a knowledge cutoff of October 2023, compared to GPT-4 Turbo’s April 2023 knowledge cutoff. Users on Plus and Enterprise plans should get access to the new model over the next two weeks, while others (including free users) might not get access to GPT-4o until after that.

What are your thoughts about where ChatGPT is heading? Hop over to our forums, where I’m hoping we can get a lively conversation going.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *


To Top