How to use DALL·E 2 API in Python - AI Photo Generation for Beginners

[@FutureSmartAI]

Thank you for the quick and to the point video.

[@1littlecoder]

Thanks Pradip for checking it out :)

[@simpilot001]

Nice video, actually. Not many that are quick to the point, speak clearly, and show every detail. Thanks

[@sithumsandaruwan]

This morning, I got two notifications. One was a gmail from DALL E saying API situation. The other one brought me here. bruh u r fast af. keep it up

[@1littlecoder]

That's so nice of you :) I made sure to make a short video before I went to sleep last night.

[@danielorlando8579]

Thanks for the video, do you have another video explaining how to use this code in Jupyter as I don't have much experience with python?

[@elementsofentertainment5493]

For some reason, PyCharm is unable to find the .Image function. What issue might this be?

[@mbegaliful]

bello! ottimo video come sempre!

[@1littlecoder]

Thank you :)

[@Jajajaja1231]

How could i diplay the image to my customer when he/she is in another website (my store website) pls answer thank u

[@1littlecoder]

I mean you can just display based on the response of the API call. Is that what you asked ?

[@Jajajaja1231]

@@1littlecoder but like what would be the procces¿¿

[@Rohit-cj6eb]

​@@1littlecoderhe is asking how to integrate this model on his website coding

[@manasjain8493]

If we use api key then also do we have any limits on how much images we can generate?

[@luxosgamer1998]

Great video! do you know how can I implement that in a webpage?

[@DiegoUceda]

Let say I use the images for the thumbnail of a video or an article, I will need to pay for each image?

[@monishmayekar5197]

Please make a video on how to download images from the url's

[@nokunato]

Hi, thanks for this. I plan to use Dall e for a image masking project. But instead of masking I want to use U2-net segmentation to select places to be masked. How can I achieve this? Thanks..

[@1littlecoder]

I think this is quite possible with Stable Diffusion, but not very sure about Dall-E, Why do you want it though?

[@nokunato]

@@1littlecoder thanks for your response. I'm working on a project which I want to launch, so I want to auto detect the mask area instead of using the masking too manually. Would appreciate an episode on it.

[@1littlecoder]

For stable diffusion, I've a tutorial to do auto masking based on keyword kzbin.info/www/bejne/mKeUiKioaKmpkJI

[@kailashprajapati5660]

Are these URLs expirable ? if yes what is the time limit for expiry?

[@mrxsal]

URL expires in 1 hour, but you can also instead get it base64 encoded so you can continue to process it with Pillow or whatever.

[@seohelpers5013]

why results are not good? panda face ... because i got this issue as well by using this API

[@Horoscope_Quotidien]

why is no one asking this question? All the images from the API are HORRIBLE. Do you have any info on this? I've tried it all and I dont understand. On bing image creatore they're amazing, same prompts..

[@sohailsheikh199]

if we put this api on our website . would our visitors have to pay the API fee or we have to pay?

[@benjiplays428]

Do you know how to integrate this to shopify website for custom prompts

[@963seeker]

thank you this was very good

[@jashbalar7526]

My account is not showing $18 which i should be granted for using OpenAI, and when i tried to run the code you described in this video, it shows this error: InvalidRequestError: Billing hard limit has been reached

[@1littlecoder]

Did you try contacting OpenAI customer support?

[@rayanbenlakdame]

in what program do i code this? i got visual studio code rn, but i guess that's not the right program for this

[@1littlecoder]

I used Jupyter Notebook inside Visual Studio code. You can even code this simply as a .py script

[@AnyReskin]

Thank you again for awesome video, btw Can you do video about API of stable diffusion too, I see the have API too to use locally.

[@1littlecoder]

I did one long ago, most of it should be relevant. Please check - kzbin.info/www/bejne/qKq1faOdec9rerM

[@AnyReskin]

@@1littlecoder thank you for answering, keep going bro

[@ThEaWsOmEgUy2011]

What that text editor called? I imagine its used in browser but could not find it

[@ThEaWsOmEgUy2011]

Another thing is that i did the code but didnt get the url for the images even if I printed the Model.list()

[@paxdii]

I got the error "openai.error.InvalidRequestError: Billing hard limit has been reached" what does that mean?

[@1littlecoder]

You get a set of API calls. You seemed to have exceeded it.

[@nik_the_pro8119]

Hey bro what program do you use for coding?

[@1littlecoder]

Bro it's Jupyter Notebook inside Visual Studio Code

[@nik_the_pro8119]

Hey when i run the openai.image.create code i doesnt output anything it just shows a green checkmark

[@diegooli9253]

same here, I should print or something to show an output?

[@Rohit-cj6eb]

Have you purchased the token for image generation?

[@ene_77]

hello i made my own colab with the dalle api, even if i change the n number of generations, and put the command print for the url why i only get 1 link?

[@1littlecoder]

That's strange. Can you put the code in a GitHub gist or share the colab with me?

[@ene_77]

@@1littlecoder for example i do: response = openai.Image.create( prompt="a white siamese cat", n=4, size="1024x1024" ) image_url = response['data'][0]['url'] print(image_url) but get only 1 link

[@1littlecoder]

Did you try for any other prompt?

[@ene_77]

@@1littlecoder yes another one and n to 4

[@lolify4433]

@@ene_77use 2 arguments if you gave 2 as n. [‘url’,’url’]

[@luis96xd]

Amazing video!

[@1littlecoder]

You are so kind. Thanks Luis!

[@ido5139]

Great video. The image quality is really poor! Why would anyone use the API if the quality is so bad

[@1littlecoder]

That was DALL-E 2, now we have DALL-E 3 but not as API yet

[@ido5139]

@@1littlecoder really looking forward for the API then :)

[@MEATHEADBooYA]

That is quite expensive if you ask me.

[@1littlecoder]

It'll be interesting to see how this entire thing plays out. I guess Enterprises might still be interested in the name tag "OpenAI" but how long that can survive with high cost and slow features.