#MrEgonAI - Simple Image Generator, Enjoy
63 messages · Page 1 of 1 (latest)
i can generate image with ratio 9:16?
yes, using gptimage you can put in the prompt
-- 9:16
ex: a horse walking into a empty street, at night , -- 9:16
result:
/based on transcription above create a prompt that would generate perfect youtube thumbnail using this template "A vibrant and professional YouTube thumbnail. A confident person in a sharp business outfit (representing an accountant), transitioning into a modern tech-savvy look (representing an automation expert). The background is split into two sections: one side shows an office with documents and calculators, and the other side features a sleek tech setup with code, graphs, and automation icons. Text overlay: 'From Accountant to Automation Expert' in bold, modern fonts using black, orange, and green colors". Replace roles using information from the transcript.
Avoid using quotation marks in prompt
thanks you 🙂
Rather satisfactory.
i get this error when trying to generate images from here
probably your browser cannot support indexedbb which is how history is saved
📝News Changelog:
** Added an (Adaptation) of In-Paint Function for GPT Image: **
- Load the image and click the "inPainting" button to open a popup with an image editing tool.
- Paint a white area where you want to edit.
- Click "Save and Submit" to send it as a new reference image.
- In the prompt, ask to edit the white part of the image to your preference.
- Other Improvements:
- Multiple Simultaneous Images (May not work perfectly):
Flux: Up to 4 images.
Turbo: Up to 4 images.
GPT Image: still 1 images.
---- Flux kontext now working
i cannot generate image by gtpimage model
gptimage is currently only working for image to image, from what was measured, it was due to the high demand of people using the model
thanks you 🙂
14/07/2025 - ChangeLog
--------------------------------------------------------------------------------
The ability to send two reference images without having to combine them has been added,
although the function to combine images still exists if needed.
--------------------------------------------------------------------------------
Now you can add two reference images and if you want to replace one of the images it is possible,
removing one of the reference images and adding another in its place,
so far I have been successful using gptimage,
kontext works but I do not guarantee such good results
--------------------------------------------------------------------------------
Combining the two images is not necessary, just upload both images, add the prompt and generate the image using Kontext or gptimage
---> 15/07/25 -Changelog - gptimage is now capable of creating images without a background, being transparent png
I am removing my project from the online status, after seeing illegal content sent as a reference image (cp) sent by people who received the service, so I am removing it from my website.
Please don't remove your project. You can validate image by ai before allowing to upload
you will work on a verification system, but while i will return to the project online again
thanks you so much
- 2025-07-23 - Change log, nsfw content selection model added and working in conjunction with reference image upload now working on the main website
the same url : https://mr-egon-ai.vercel.app/ working again
thanks you, hehe
you can share the source with us |?
How can I get more followers and subscribers
**Changelog / 07/30/2025: The image host has been changed, to now be more operational, most likely causing fewer problems in image models like kontext **
When I entered the site, it frooze my device for 4 seconds and also site crashed
It is not so good for mobile, as it uses the indexed db so that the history remains private for each browser, which ends up causing performance problems for mobile devices.
Also another question, How does image to image generation work?
You wait for the site to load and upload your reference image using the reference image function, there you use models such as (gptimage - currently unavailable) or the context to create edits to this image, that is, you upload, select the kontext model, make a prompt based on what you want to edit, uncheck the custom size option, and generate a set of images 
for me the site is in my language, but the concept is the same 
I have to give ai UIL of images?
How do you host images?
There are several sites out there that host images via API key for free
I just make the call to the API to upload an use the reference image
Can you seggest me some of em?
imgbb is one
Provide free and unlimited?
image upload yes, but for calls via API key it is limited but has a lot of use, but after exceeding you can generate a new key and use it
Which one do you use,
I used to use it a lot, but I don't use it anymore,
Then, which one do you use now.
this one, freeimage.host/
This is what, so far, has made me not have so many errors when using the reference image + the flux kontext
Thank you. Is gptimage avaible via pollination.ai now?
hey bro where do u upload your app?
@supple sorrel
@hollow oracle github
ty
nanobanana released? of course i will insert