OpenAI rolls out new GPT-4o model in ChatGPT, boosting multi-step reasoning
Share link:In this post: OpenAI’s GPT-4o now excels in multi-step reasoning and independent image generation. Users notice enhanced accuracy and detail in the new GPT-4o model in ChatGPT. Transparency concerns rise as users question the differences in the latest GPT-4o update.
In a recent update to the ChatGPT model, OpenAI has launched the GPT-4o variant that brings with it significant improvements to multi-step reasoning and detailed explanations. The official ChatGPT account quietly announced on social platform X, but users quickly picked up on differences in the model’s performance.
ChatGPT’s new GPT-4o model, which is already live within the system, responds with more comprehensive step-by-step reasoning — a change many users have found helpful. This improvement has led to higher accuracy responses, particularly for complex queries that require logical flow and detailed explanation. Users at large have also noted stronger inherent image generation capabilities resulting in higher quality and accurate visuals compared to its predecessor.
GPT-4o takes over image generation, leaving DALL-E 3 behind
The most notable modification in the updated GPT-4o model is its improved ability to generate images. Formerly, OpenAI’s ChatGPT relied on DALL-E 3 to create images from text prompts. However, this has changed with the latest release of GPT-4o which can now generate images by itself without using DALL-E 3. This is significant because GPT-4o was designed natively multimodal and capable of treating both text and pixels as tokens. Therefore it can create better quality pictures faster and understand text prompts better.
The ability to generate images independently speeds up and improves the accuracy and reality of generated content. Users claim that visual outputs by GPT-4o are more consistent with their input texts thus making the whole experience seamless and integrated within ChatGPT environment.
Users demand OpenAI’s clarity on GPT-4o update changes
However, not all feedback has been positive. Specifically, users have criticized OpenAI’s lack of transparency about what exactly changed in this update. The company has yet to fully describe how this version of GPT-4o differs from earlier ones – leaving many people to wonder about distinctions between the ChatGPT variant and the version available through OpenAI’s API.
Acknowledging these concerns, an OpenAI spokesperson said that indeed there may be differences between the ChatGPT variant and the API variant. This is because they are still working on making the model better suited for different use cases; while optimizing it to promote user interaction within ChatGPT but focusing more on developer needs with the API version.
With increasing numbers of people trying out updated GPT-4o, there will be a growing demand for additional explanations from Open AI. At present, however, chat GPT improvements are being made to improve user experience but questions about what exactly has changed remain unanswered.
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
Do Kwon’s extradition case goes back to Montenegrin justice minister
SEC seeks sanctions against Elon Musk over Twitter purchase
SEC greenlights listing and trading options for BlackRock's spot bitcoin ETF
The SEC approved spot bitcoin ETFs earlier this year, but had not yet approved listing and trading for options on those products.
Securitize adds Wormhole for multi-chain RWA tokens