Why does DALL·E lag behind its competition?

Dear OpenAI Team,

I appreciate all the advancements you’ve made with ChatGPT and AI in general, but I’d like to express my concerns regarding DALL·E and image interpretation capabilities.

  1. DALL·E Improvement – Compared to competitors like Midjourney and Stable Diffusion XL, DALL·E 3 feels outdated. The image quality, realism, and consistency fall behind industry standards. It would be great to see a major update that improves fidelity, creativity, and coherence in image generation, aligning with OpenAI’s high standards.
  2. Image Input for ChatGPT API – It’s quite frustrating that, despite ChatGPT being able to interpret images in the web app, the API still lacks this functionality. Many developers and businesses would greatly benefit from this feature being available. Are there any plans to enable image input in the API soon?
  3. No Image Support for More Powerful Models – It’s disappointing that the most advanced models, like GPT-4 Turbo (o1), still do not have the ability to process images. This severely limits their usability for many real-world applications that require multimodal understanding. OpenAI is known for pushing the boundaries of AI, so it’s surprising to see this limitation persist.

I hope OpenAI will address these issues soon, as they are crucial for keeping up with the rapid advancements in AI-generated content. Looking forward to seeing improvements in these areas!

2 Likes

This is what I filed a report to OpenAI a few weeks back without any reply: An ongoing issue on chatpgt is that the reasoning models like 01 have been released for a while now and since OpenAI had directly integrated the Chatgpt4 model with DALL-E, I assumed they would also not wait to long with integrating it further into the updated new models that are able to reason because they are already announcing that model o3 now in Februari 2025 which is very close, but then on the other side they have yet to link the o1 with DALL-E while there is already an o3 almost being released and that is some incongruent product release to an also pretty important product (DALL-E) which ironically was an important part in the growth of AI. Seeing how fast they can build and release very advanced technology I would assume making this Update in the Reasoning Capabilities of DALL-E is not too difficult and can be done pretty easy and fast, but they still haven’t DALLE 3 is still not able to comprehend what you are exactly asking even if you make the prompt accurately specific, on top of that you also cannot further ask to adjust the Artwork as it will create an entirely new prompt each time, which is very frustrating. DALLE 3 is seriously lagging behind and for a multi billion company that is just not done after 2 years almost no improvement.