#community-help
1 messages · Page 9 of 1
man this is why we can't we have nice things
Is it normal that after using outpainting the noise is different from the original dalle output?
Yes. Happens to most of my images
I'd show you but I used the blur tool in photoshop to fix them all
Ah except this one
I used denoiser after selecting the noisier square and using 16px feather. looks good enough now
hey there !
I'm not a 100% sure it was Dall-E but I tired once a few months ago a system of a discord Bot that generate image.
And when we had the 4 generated picture, we could ask for variations, ok, but we also could ask the bot to work more on the selected picture 😮
Why can't we do it anymore ? 😦
The system you’re referring to is known as MidJourney. It’s not affiliated with DALL E
Oh ok thanks 🙂
I was kinda lost x)
So I found a prompt that consistently generates images of very similar looking girls so I'm a bit concerned they are deepfakes what do I do?
Report it to support@openai.com
Do i lose renders if I dont save them? I made some really nice images about a week ago but now I cant find them in my recent library. Is there any way I can locate or find a render from last week?
i think so, not 100% sure tho
Browser history should still have everything
I'm curious if this is a case of the filter being over zealous or if there's anything in this prompt that would be against the rules?
Could be a few things. "Red skin" could be seen as a circumventing measure to produce blood, especially with "ripped" and "spear." The filter considers context.
Hi, I just recently made my account with google and I was wondering if there is a way to set a password so I can access my account from places where I am unable to sign into google.
Good question. AFAIK once your account is tied to Google, it's that way for good. But the place to ask is support@openai.com
Thanks for the reply. Yeah my organization disables the access of google accounts on the corporate network. I should have thought of that :S
I created an account with a work email last week. Is it possible to invite others in my company to use dall-e under the same account?
No. They can use your account under your supervision. I'd suggest you direct them to the waitlist, as the turnover on new accounts seems to be pretty fast these days.
Thanks
If I did login with google, can I convert to username/password so that I can share that?
no (see the above comment) and by supervision, i mean DIRECT supervision. your account is tied to YOUR agreement of ToC and content policy, and every generation is tied to you personally
Hi! Does anyone know if i can prompt in spanish?
you definitely can! mileage varies on non-english languages though. it works best for subjects that would have more data in that language than in english (e.g. culturally specific objects or ideas)
That’s great! Thanks for the info!
Someone explain to me how I could upload an image and ask him to invent the continuation?
Hi, are you referring to outpainting / uncropping? If so, see the bot's message below for some info and steps.
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
Thanks a lot
@verbal crag Permission to DM regarding a matter, I believe is best not asked here please...
Hi there, was hoping for some clarification when it comes to uploading photos of real individuals.
I'm working on a project based around early photography in non European cultures and have not been able to get DALL-E to match the intended results as closely as I would like. I have a photograph of an individual that was taken in the mid 1800s and wanted to know if that would still violate the usage policy of uploading someone's likeness to the bot.
Thanks in advance for any clarification on this issue!
As a slight answer until someone more official comes, I'd imagine it would still be against the policy because they have no internal systems to verify the image is from the 1800s
It will probably trip the content filter regardless of when it was taken. The policy is pretty unambiguous, and while your intent is not harmful, the protections are an umbrella for safety reasons and I'd say your use case here is not within policy.
Hey guys, I see a lot of people triggering the filter but no one thinks much of it, so today I accidentally triggered it while trying to recreate a prompt I saw people used but it got flagged as going against content policy and that further violations may get my account suspended. So i wanted to check with you and make sure I'm not risking to lose my account(the prompt was a funko pop of a famous character, nothing gory or violent)
I would not recommend trying to do something that actually is breaking the content rules, but a single accidental trip will be ok.
• Our filter aims to detect generated text that could be sensitive or unsafe. The filter will make mistakes and we have currently built it to err on the side of caution, thus, resulting in more false positives. We're working on improving our filters, so this should become less of an issue in the future.
• While we're not sharing the exact thresholds, a few warnings here and there won't lead to an account suspension.
Yeah I was just curious because someone else made it work(albeit with a different character) so I panicked for a second.
Thanks for the answer!
Hello, do we have to attribute every image generated with Dall-E 2 to Dall-E 2?
The latest communication from OpenAI is this: Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL-E 2. We'll update the content policy to make this clearer in the coming days!
Hi! Does anyone know if we can get different aspect ratios or just the standard square images?
just square at this point in the beta 🙂
but you can use outpainting to make an aspect ratio you want. see #inpaint-outpaint
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
How can I send someone an invitation to Dall-e?
Currently, you can only gain access to DALL E 2 through the waitlist.
Thanks
If i'm editing a part of an imagine, do I have to describle the whole image or just the part I want to edit?
can i generate images here? by command??
No
At it's simplest, Inpainting is a way to replace parts of an image using DALL∙E's edit feature.
(There's more you can do with inpainting, this is the most basic and commonly used. Check the #edit-inpainting channel for more ways to inpaint.)
Upload an image or select one that you've generated with DALL∙E.
Press Edit, and erase the portions of the image you want to replace.
Enter a prompt describing the image as a whole (in the example, the prompt changed from "A cat sitting on a bench" to "A dog sitting on a bench") and press Generate.
@gaunt sail see above
whats the best way to go about submitting things to be added to the dictionairi?
#❌┃tips-n-tricks, make sure to mention @rugged wing
am I allowed to use an image I generated as a logo for a company I'm starting?
Yes.
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
thank you!!!
Any time 🙂 do read the content policy and terms. But you are good to go.
I have a question if "do not attempt to create images of public figures (including celebrities)" guideline also applies to fictional characters? Or who can I talk to about a specific prompt just in case to not generate something unwanted in beta testing?
You can create fictional characters. But it may or may not work well.
Thanks. This character has been rendered artistically and played by actors many times so there is enough data, but different engines do different takes on the prompt.
"But it may or not work well"... Well yes I was surprised! But it worked very well by image quality. This is a good test subject I think. Disco Diffusion is the only one that had no luck with this.
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
ty
if you dont use credits do they keep on stacking every month past 50?
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
will my first free 50 credits disappear next month?
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
oh so i gotta use my 50 credits
• While model updates aren't frequent, we can definitely consider adding licensable data for it in the next training round.
• Please tell us more about it through this form.
Please treat others the way you would like to be treated, and assume best intentions. Don’t harass or attack others, and don’t engage in hateful or generally malicious behavior (e.g. sexism, racism, homophobia, etc.) Community volunteers will issue warnings and then mute, kick, or ban you from the server — depending on the severity.
Hi when i go to labs.openai.com and use the edit feature it’s just gives a blank page
Try using another browser and see if you get the same issue
Hi, i uploaded a few music album covers to get variations of, does that go against the "Do not upload images to which you do not hold appropriate usage rights." rule? I genuinely did not see said rule until now since i got paranoid lol
I'm unsure of the answer here, I'd say just do not post them anywhere :)
Ah ok, thank you
I wont at this point just to be safe
On the same topic, would it be ok to upload like, art? Like if someone uploaded Starry Night or something to get variations of
Sure :) totally fine
Ah ok thats nice to know
Are uploaded images stored/saved?
yes, they must be, as they remain visible as a thumbnail "original"
Is there any way to propose fiends for fast track? 🙂 I think I waited at least a month on the list myself
No, that's not possible since there's no 'fast track' 
The turnover on new accounts seems to be pretty fast these days though!
Was just wish full thinking 🙂
Can I use Dall-E 2 creations as a reference for the digital art I do? I post on Instagram, but should/can I post the original AI creations next to it and credit Dall-E 2? 😊
Sure!
Can I use my Creations for everything I want? like if I publish music, using it as artwork/cover art?
I think you have to buy the rights for it unless u credit Dall-E 2 on the cover or something, but I'm a bit unsure!
Yes.
The bot is the true MVP
Thanks!
Hello Guys! How does Dall E 2 image command works in discord?
There are a few custom functions in #dall-e-bot, but DALL·E operates through labs.openai.com
Thanks jeremy-o 🙏 so there is no comand like /imagine like midhourney
correct.
appreciated your time
whats the idea behind the credit system? I mean, why not subscription based?
a subscription at this point would still necessarily have a limitation on credits because of the number of users in the beta. OpenAI need to understand use patterns as they expand to a million beta participants to protect their resources from overload
hmm, alright. thank you.
Any good way to remove the watermark?
In your results page, you can expand an image and use inspect element to grab the unwatermarked image url. The domain it will be on ends with windows.net. It's a webp so you might want to convert it afterwards.
Thanks, that worked!
Is anyone else locked out at the moment? Logging in just takes me back to the 'invite required' page. I am paid up too.
Its been all day and I cant log in. HELP
Had this on mobile yesterday. had to give it like 4-5 tries to get in
On PC I havnt got this problem yet
Im on my normal pc. Tried dif browsers. Is there no official support thread ? Tried resetting password, which worked, but the new password doesnt get me past the 'you need an invite' redirect. Arghh
Not having any issues personally.
You can contact support@openai.com
Thanks - sending support email now
I forgot that I applied in with another email I long time ago... I haven't touched it or opened just in case, does that get you banned?
Is it okay to post my creations on instagram? As long as i state they are Dalle2 images right?
Are there any pictures I am not allowed to generate with certain words, cause they would violate the terms?
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
Yes, your can post your images on social media
noticed that files created are about 1mb and 13x13 at 72 dpi on average. is that the max? @dull basin
[❓] — Q: can I create images of higher resolutions than the standard ones?
Upon trying to add a payment method:
We cannot currently accept payments from Ukraine
How come?
How do you all manage with the credits and stuff... I need DALL-E everyday for sanity. haha I'm getting through my tries far too quickly lol
Will you ever make a subscription tier?
Currently DALL E can only generate images at 1024x1024. However you may want to take a look at patch-e.com, it may facilitate a little of what you’re looking for.
As long as you’re not attempting to abuse the system, I don’t think so. I wouldn’t recommend exploiting the free credits you get on your second account, personally.
From OpenAI:
I have been trying to create a picture that it seams is not possible for the AI ?? its of "a fox with nine tails a forest in the background, many separate tails"
the Nine tails have never showen ?
Try #❌┃prompt-help . DALL E often tries to rectify what is in the prompt with what is in its training data. Which means if there’s lots of foxes in the training data but all of them have only one tail, it will most likely produce a fox with only one tail.
I don’t have an answer for this at this time. I can reach out to the OpenAI team and try to get some more information.
Can i push it to focus on it ? i did try the edit and mark the tails and type many tails + the name of an anime that i know have that.
I would appreciate that, thank you 🙂
#❌┃prompt-help is your best bet for working on this.
Okay thanks for you time and explanation 🙂
I’m waiting for an official answer. However, it appears this may be due to the payment processor: https://support.stripe.com/questions/impact-of-sanctions-on-russia-and-belarus
Find help and support for Stripe. Our support center provides answers on all types of situations, including account information, charges and refunds, and subscriptions information. Get your questions answered and find international support for Stripe.
hey i was wondering, is there a dalle 2 app?
No app, just the web interface at the moment. If you're on mobile, your device may let you "add website to home screen", which will create a shortcut icon to labs.openai.com that acts similarly to a mobile app
Hi, thank u for creating the amazing Dall-E 2 tool. I really love using it. I'm trying to edit this picture I created with Midjourney, but the AI tells me it may violate policies and that further violations may lead to my account being suspended. I read all the rules of course and can't see how can this picture violate your policies. Can u please clarify and help me understand what the problem could be?
@lime gazelle to prohibit the use of personal photos showing a face they have banned the upload for all pictures including a face
Any kind of face, even clearly digitally drawn or AI artistically drawn and not photorealistic?
But I'm sure I can upload pictures with faces created by Dall-E, because I use it to create illustrations and then outcrop them in Dall-E after re-uploading the pictures.
How can this be explained?
In any case, if I remove with photoshop the face before upload, then it'll be ok in your opinion?
Are there any plans on adding another payment method?
@lime gazelle
Q: Why can’t I upload my own face?
• To prevent deepfakes, we are currently prohibiting uploads of all realistic faces, even when the face belongs to you or if you have consent.
• This is because we don't yet have a scalable way to verify whether you have consent, or whether a face belongs to you. So we prohibited all face uploads.
im an newby but sounds fine to me
As far as I know there are no news regarding adding new/different payment options.
Some people have had luck with Revolut if that's an option for you 
How do i change too 16:9?
Please refer to this answer 
#community-help message
@pine void Okay thanks 🙂
if i don't use all my free credits by the end of the month, are they wasted?
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
Have a look at the message above :)
Dall•E might not even be bothered by the face; instead, it blocks you from
uploading the image because it recognizes a firearm in the picture.
Have a read through the Content Policy to understand what
is and is not allowed 
https://labs.openai.com/policies/content-policy
Meaning we cannot draw cyberpunk weapons with Dall-E or can't upload them?
You can't generate and/or upload them :)
Can I ask something with blood on?
So I need to apply vfx to a shot. Can dall-e be used to generate gun shot wounds or is that against policy?
yea kinda same question
hahahahah crazy that thats the exact question i asked hahahaha
what are the odds
jinx
lmao
imagine the capabilities and time saving if we could do this
one head 4 hands lmao
GAHAHAHAHAHA
Does anyone know why it doesn’t render faces of well known, famous people? I noticed this with several inputs.
Since fake blood is so similar to real blood, it violates the Content Policy just like real blood does.
@timber epoch This should answere your question too 
dal-e-bot query. I'm probably being stupid but I can't see how to load a picture that has been padded by dal-e-bot back into dal e for processing ?
missed opportunity. they will need to figure something out for vfx artitst in future.
maybe a unique with everyone who generates such and a log of what they generated so it can be traced if it is used for anything wrong
unique code
Please feel free to submit any recommendations you may have in #suggestions :)
When you click the button upload an image, you should be able to upload your padded image 
I've been searching for an answer and not finding a clear answer; if I upload an image, can I have dalle2 adjsut the medium of the image or just sections? Didn't want to burn through my credits without asking first.
IE. upload screencap of a game character, have Dalle2 paint that character?
how does one go about expanding upon the square aspect ratio of a creation?
Why does the Dall-e bot crop out a portion of the sides of an already composited image, when trying to do another composite?
Like here
There is no dedicated style/medium transfer feature at the moment, although some have found success via inpainting. More details: #openai-chatter message
There are plenty resources for expanding aspect ratio through compositing images 🙂 Some are linked in #gpt-models. Additional links: #community-help message
One popular website for this is https://patch-e.com - accompanied by a 10 minute tutorial on the process: https://youtu.be/ZQHEmYt8g3k
Will look into that
but i guess that the limit amount of times that you can do compositing with the discord bot, deriving from an original image, is 2 times, after that it doesn't increase it's size horizontaly
I just signed up so I have to rush to use my 50 free credits or do I have till sep 28th?
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
I think it says that it will expire on the 28th?
Correct.
On the "What would you like to teach DALL·E?" form, on the third question, are wikipedia gallery links ok to use?
I'm not at all familiar with "datasets" to find out where they can be to help train Dall-E on a certain subject, so...
How do I get the faces more clear? https://labs.openai.com/s/b4DFolnSEmgEh70VEj8y3HBe
what did I do wrong?
KAWS is a person/celebrity? If so then it's going to be blocked
is DALL-E CC0, are we able to use images freely or no?
an artist so I thought it would be in the style
If you want a style try saying "by KAWS" @lean delta
Two faces in a very wide shot are unlikely to become good.
You best option is to generate a close up or portrait and expand that.
Okay thanks
probably a common question, but can I get back to older history items? possibly by deleting more recent ones somehow?
For generations past the most recent 50, you'll need to have saved the unique labs.openai.com/e/ address for each prompt you ran. These will be in your browser history if you have it enabled
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
oh man 😔
rubber is probably triggering the content filter
Ah, yeah I tried my suggestion and it still gave a warning
that's frustrating
I was just sitting here trying to figure out what on earth I was doing wrong with rubber ducks. Was worried I had been blocked but not locked out yet when first I had asked for "pirate rubber ducky on the high seas", but even when asking for "rubber ducky on the open ocean" it is a no go, so I'll bet you it is indeed "rubber".
Is it possible to combine two sample images into one, and see what the AI generates?
You can do this:
Is that what you mean?
Thanks you
uhh, making dall-e pfps are legal? if so what is !watermark in here?
Yes. !watermark was developed and released way before beta, we decided to keep it for people who'd like to watermark their PFPs.
so its getting copyrighted to make no one use i mean if so how do we report it?
When I go into "my collection" It isnt showing any of my generations, is this a bug?
have you saved any to your collection?
When I created image with verification code it still used a credit, what should I do about that, is it possible to get that credit back?
Unfortunately, no. As it's really supposed to use a credit. Here's an excerpt from the #verify channel:
The bot will ask you to incorporate a numeric code into your next DALL·E prompt, using a DALL·E credit -- and the code won't affect the results (so you wouldn't be "wasting" a credit).
The idea for this is you will use a credit anyway, adding a code in the does not affect the generated results.
Oh, sorry then, aš english iš not my native language, į thought that it iš written that credit will not be wasted, sorry
No worries, welcome to the server! 
If we have artist subsidy, do the credits remain after the month is over as paid credit or do i need to use up all August credits straight away?
Hi, please check your email regarding that. I believe they've specified how Artist Subsidy credits work there.
Hello, I want to try different art styles. How can I tell the AI? I could not find anything where something like this was written down
There are guides over at #gpt-models 🙂
@fossil lintel — hey, for now you can save your collection one by one. You can click this icon to download your preferred image.
oh so just one by one then... I thought i can save the whole generated collection and have it in my DALL-E account. Ok thank you
It doesn't seem to know many fantasy creatures or myths... near impossible to get something shaped like a centaur and good luck making a cyclops...
@verbal crag is this possible for DALL-E to create a four-piece wall art, like that wall art in the picture?
Totally!
No automatic way though, you'd have to manually generate all four tiles and manually stitch them up.
Some have been recommending a website called patch-e (patch-e.com).
Check #inpaint-outpaint for examples.
Most likely been asked before, but is there a timeline for releasing an API for applications?
are you allowed to sell a picture generated from using dall-e? (assuming the source is referenced perhaps?{)
DALL·E is not yet available through an API, and users are permitted to access DALL·E exclusively via our public-facing website. We do not permit programmatic access via web scraping or data extraction, as outlined in our Terms of Use. We will continue to evaluate how we release this technology as we learn more during the beta.
No timeline released yet. Stay tuned for updates.
Yes.
@undone plume — you may refer to the terms in #old-server-rules under Rule 0.
thank you. i was going through the wrong bit
What command should I use to resize the picture to 16:9?
Please refer to these two answers 
#community-help message
#community-help message
Do commands used in the dall-e-bot room affect our credits?
Hi! I just wanted to ask, I'm a musician and i wanted to ask if it is allowed for me to use some of the ai generated artwork as cover art. I will likely make a few changes or touch ups to the images but i wanted to know if it would be ok for me to use them as well as the terms associated with usage
You have full commercial rights to your images
Oh i see, are there any terms/ conditions associated with usage of the images?
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
ah yes, thank you!
Hey, new to the group ! Can I create images with different formats ? Like 16:9, 3:5, 5:4 or it only create squares ?
Hello! I'm not so good at understanding all those terms and policy's but are we allowed to use the images we create with Dalle to print on canvases like mugs or tshirts for own use and secondly is it allowed to make money from the pictures you make like putting the print up on merh sites or is all right to the pictures on their side?
All I could find/understand and idk if it's right but all sharing of the pictures should somehow link/be described to be made by Dalle?
Sorry if I sound fuzzy
Please refer to these two answers 
#community-help message
#community-help message
Yes, as long as you follow the policies, you have full commercial rights to your generation(s) 
More details can be found in #old-server-rules, under Rule 0.
How would i go about getting art in these styles
Looong descriptions and something like "digital art" on the end usually works for me
Are monthly credits stackable?
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
Could u like give me the prompt and i try it out
Is there a way to change the aspect ratio
In its current version, DALL·E can only produce images in a 1024x1024 square.
Please refer to these two answers 
#community-help message
#community-help message
how do you set a specific color of any element?
What is the exact problem?
DALL-E is not that with colors. Writing "red X" will often make X red, but it will often happen that if you say "red X" the whole picture will have red elements.
Hi everyone, looking for a kind advise from the community please - how often are we encouraged to 'Report issue' when the results doesn't match the prompt? I.e. if I specify in the prompt "dim light from two bulbs on the wall" and all generations have light sources on the ceiling?.. (please don't get me wrong - my motive is purely the model improvement). Thanks! 🙂
For example, I wrote "... golden clouds ..." but now I want to edit the places where they are too bright, so is it possible to select the areas and write something like "darker golden clouds" for it to produce same results, but different color?
Looking for advice for when DALL-E 2 will or won't flag something as a TOS violation. Hypothetical examples: 1) Text includes the name of a person who is deceased (e.g. "Abraham Lincoln"). 2) Text includes the name of a current entity named for a person now deceased (e.g. "Lincoln Memorial" or "Bryce Canyon"). 3) Text includes the current name of a building that belonged to a person now deceased (e.g. "Abraham Lincoln's Cabin"). I got cited for something similar to #3, and am trying to make sure I don't make similar errors. TIA! 🙂
That will be hard. It might work if you say "dark golden clouds", but you won't have exact control.
Hi everyone!
I generated a picture with DALL-E but I forgot to save it into my collection and got rolled out of Recents. I have a copy of it, but I want it in my collection. Is there any way to get back an older picture and put it into my collection?
Try using your browser history
The exact guidelines for people who are and aren't allowed to be generated are still being finalized. Hold tight!
Where do the files save to when I download an image? I’m using an iPhone and there’s no prompt for me to choose a location.
It worked!!! Thank you very much!!!
Never mind. I determined that the issue was the browser I was using.
Hopefully a quick question. On the labs site I attempted to upload an image of something I created per the content policy, I own all rights. The system flagged me for a violation though it does not seem to conflict with any of the other polices.
- I guess the question is, is there a process or step that I've missed so that I can upload something I've created?
There are some policies specific to uploads.
• Do not upload images of people without their consent, including public figures.
• Do not upload images to which you do not hold appropriate usage rights.
• To prevent deepfakes, we are currently prohibiting uploads of all realistic faces, even when the face belongs to you or if you have consent.
@split hazelI read those on the site.. that is why I am confused, I created the image (by hand from my imagination) I think that means I own it, I would think that means I have the appropriate usage rights. The face is not realistic but there is a drawn face. Did it flag me because the image had a face in general and cannot determine if it is realistic or of a celeb or something? just cursious
On rare occasions it will falsely detect some things a real face, that's just an inevitability atm 😦 I'm not sure how close to realistic your drawn face is but if I had to hazard a guess, I'd say it's the cause. It would help to have reference to the image you're referring to if you'd like to pursue further
@split hazel thanks for the info. If I try again with my art all just try not having a face. I just didn't want to get banned or anything.
I can send you a dm of the image I was playing with if you are curious
Sure 
@split hazel sent it.
Why can't you enter "Michael Jackson" or "Justin Bieber", for example? It appears in the search as a content against the guidelines.
The content policy forbids the upload or generation of real people
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
can the dalle discord bot not pad horizontal images?
or did i do something incorrectly with my attempt in #dall-e-bot
The discord bot currently only outputs square images, even when an uploaded image is horizontal/vertical.
Looking at your attempt, try a bigger number, like 50 or so.
will do! thanks for the replies
#❌┃tips-n-tricks message
@vivid mortar do you still need to feed a prompt when you edit the image
Yep! Describe the entire image you want to see
since dall e 2 cant generate prompts with copyrighted material since its in the rules, may i ask as to how its possible for the openai team to get in trouble for an ai generated image?
(i hope i worded this right lol)
people can be creative i suppose
i dont need to tell it to create danny devito for it to accidentally (or intentionally on my part) create something in his likeness
ive seen a bunch of people "zoom out" of album covers using Dall-E 2, how do they do that?
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
ah thanks DALL-E bot
You're welcome!
hi, are images we create in our account private unless shared using the 'share' button? or are they published via some bot to a channel/site? (like happens with free midjourney account)
Hi, all your images are private until you publish them. After publishing an image it is only accessible to those with its /s/ address
Is there a central list of banned words? Such as "shot" (for those of us thinking of photography) and "rubber" (for those of us imagining rubber ducks) So we can figure this out a little quicker?
I was told there is not
Thank you
No, there is currently no list of banned words.
However, you can always read the Content policy to get
a better understanding of what is and is not allowed 
https://labs.openai.com/policies/content-policy
Do historical people count as "public figures" like Marco Polo for example?
and
Does history count as politics?
and another one, do fictional characters count as public figures? i.e saul goodman or superman
No you are free to do those
thank you for the answe 👍
Have a read through the
Content Policy: https://labs.openai.com/policies/content-policy
Historical figures are typically acceptable, some however are blocked.
(Such as dictators etc.)
yea, i was more thinking about ppl like alexander the great
thanks for the answer
Thank you
is there an API you can use yet for automating image generation? (of course still using tokens and some kind of authentication)
It is a bit awkward to ask 😅
But is there any reason why my image has 11 ⭐ and is not in #hall-of-fame
good question! @verbal crag might know - i believe there was a timeframe between bots when it could have potentially been skipped over, but i thought that time was over with
unless i misunderstood the change that was going on
could also just be an error with what the star bot is set to monitor?
Hi is it possible to add arguments for aspect ratio for dalle 2?
someone stealed my pfp with watermark can i do a thing?
not at this time
There is no public api at this time
sad :/
uhh
Gotcha 
You can let them know you retain the rights to your generated images.
alright but he blocked me
Thanks for reporting! Apologies as I've overlooked a setting, I have since fixed the issue. Great photo by the way!
cc: @past geyser
Are they a member of this server? If not there really isn’t anything you can do without legal action. Every online artist has this problem
Thank you so much! 
hey gm - can I use the dall.e pictures for my website?
Hi, is there any way to move a friend up in the queue to get full access to dall•e? He would like to introduce it to his company on a meeting asap, but he registered just now
Yep!
Yes.
that is realy great, thx
Hi! At the moment, no. No worries though, we've seen the wait get shorter and shorter.
Okay, then i have one more question for which i couldnt find the answer in the terms of use. Is account sharing prohibited?
You may check the API Terms & Policies (OpenAI's Terms of Use) in #old-server-rules, under Rule 0 for that.
This may expedite your friend's access though, no guarantee: https://openai.com/contact-sales/
Should I submit the form or does he need to do that himself?
It would be better if he submits the form himself.
Does boosting the server (even though it's far beyond level three) grant anything other than the role?
I have two boosts on a server that started with 0 and now I have got people boosting past the threshold so one of my boosts isn't needed any more. 
Server boosting gives the server boosters role, which also grants access to a boosters channel 
OK spicy
So the one ambiguity I would like cleared up is: How do historical figures fit into the political figures/celebrities thing?
Like, is FDR ok? Marilyn Monroe?
How far back do you have to go?
.
@rose glen
Sorry if this is a dumb question. Dalle was NOT making such good faces like this one month ago. Has it learned?
the model hasn't changed - it's static. as a user since june, i can assure you it has been making such good faces for more than a month! it's just not always consistent with its output, so it can seem like it changes
e.g. this was created on june 23rd: https://labs.openai.com/s/nm1WaCGOrZpFHxtuRH2JKvtP
I'll chime in, these were created early July
Ok amazing. Thanks for responses. This is what I was getting in July! But I understand now that prompt matters. https://labs.openai.com/s/nlJFR9aI6AXrpSGRQbyyir6N
I’m sorry but please never ever use that prompt again
It was for a particular project!
What was it?
In your case, DALL-E 2 will have difficulty generating multiple faces.
Best to generate just one face per generation. Then inpaint the rest.
You can actually do two faces at ones if you take a often photographed motive like couple photography.
Interesting. Will have to try that sometime.
See here for a few examples and a prompt.
A film clip for a song about the party we'll have when the world is about to end
Did you have party after the world ended cause that explain the faces 💀
Yeah exactly!
So uploading real photos of humans isn't allowed. What about uploading a face of human that got rendered in, for example, Blender?
Something that is recognized as a face will be blocked
You can try! Although it might get blocked if deemed too realistic :(
They look always cartoony and never like real human (probably because of missing normal mapping)
hi! apologies in advance for the ridiculous Q, but... on midjourney you can message the bot with the prompt /imagine and get your images on disocord. I've been using Dall.e 2 on the browser, but can't figure out how to create images using the bot on discord. what am I'm missing? Thanks for taking the time to reply 🙂
You can only generate images through the website, labs.openai.com
ahhh that makes sense. all good on that front. thank you
if I get a bunch of "image doesn't match my text description" can I request credit refunds? these are so wrong i'm sure there must be a crossed line. Or, noting their wrongness is paramount to making this work well
I don’t believe they issue refunds for prompts that don’t match your descriptions. I highly recommend checking out #gpt-models and reading the guides there.
you realize my situation tho? i'm using the report button per the guide
not just unhappy with results
You may try contacting support@openai.com but no guarantees.
reasonable. ty
Very polite of you, is it reasonable to have to email a support address to get a refund for a credit that generated nothing even close to your prompt and there are "no guarantees"

Unfortunately, we can't guarantee anything here as we're only volunteers. 😦
I'm not blaming anyone here
So it's up to the staff at the support@ team to handle these kind of situations.
can you add paypal as a way to buy credits. I don't like saving my card info on sites.
The current payment system is the first. There may be changes to come, but we don't have details right now
i personally use privacy to avoid worrying about this. great for a lot of sites. it's like a credit card proxy. no cost. they likely make $ from data. this is my referral link in case you're interested ^_^ https://privacy.com/join/D7TRE
Also, can prolly pick up a Visa/Mastercard giftcard from the store and load it then just maintain that for a method of using a throwaway
unfortunately my bank doesn't allow this for some reason, or they don't allow my bank, whichever it is. I tried various CC website like such. No luck. Thinking I should grab a pre-paid at the store lol
can you upload lewd pics to make variations without actually sharing them? does that go against the policy?
You cannot, it does go against the policy
even for my own use?
correct
yes
good to know
anyone else having issues with "upload & generate"? keep getting "Something went wrong. Please try again in a few minutes. If the issue persists after trying again, please contact support@openai.com."
Im getting that as well now , can someone help?
Dall-E is currently having some server issues. Give it some time and you should be able to generate and/or upload images again 
Check my message above :)
i see, thank you! will try again later 🦦
Okay I tried multiple times today to buy credits (It didn't work because our bank is slow)
Now it says that my card has been declined to often - does that mean that it's unusable now or can I try it in 24 hours again?
As of right now, it seems that there isn't enough money on it to buy credits. What if that changes in probably 24 hours?
Or das that mean that I can't buy any credits anymore with my current card? That would be horrible
i would email support@openai.com
I have the hope that it will work in 24 to 48 hours.
I have enough money to buy credits, I did load it to the card but it never worked today.
I removed the payment method for now. I'll try it in 24 hours again.
Does the server still has problems? I haven't been able to create nothing the last 20 min
yes, watch #announcements for updates or subscribe https://status.openai.com/
Welcome to OpenAI's home for real-time and historical data on system performance.
they've identified the cause, hopefully be back soon - fingers crossed
Thanks Alex
I did that, now I just have to wait
good luck. i hope it gets fixed quickly for you
FYI: Update - We have moved text-ada-001 and classic davinci out of our affected cluster. These two models are now operational. Most DALL·E traffic is still heavily degraded.
Thanks
I hope so
Does anyone know of a program that can turn 2D objects into 3D
Basically trying to do that with these
I think photoshop has some basic functionality like that
is there a good list somewhere of digital artists dalle recognizes?
particularly any with a pink, feminine, soft aesthetic?
non digital artists are welcome too!
Sorry if this has already been answered but is there any process to get an image I generated approved for commercial use?
Is it possible to extend the crop of an image? I have a 4:3 image that I want extended to 21:9
That's about as good as it gets
Yes, you have full commercial rights to your images
Try patch-e.com
I managed to do it with Dall-e, by putting in the empty space as transparent around the image
nice!
Thanks!
it has some ugly lines at the old crop border though, I did not expect that
Yeah, it does that 😦
If you erase it in a jagged pattern it blends it together much better
Yeah, I did that on my second try
These look like mario kart levels lol I love it
still a bit of eeeeeh but it works better
thanks
Output from MidJourney's new beta that I try to clean up and wide-ify with Dall-E
Very cool! What were your prompts and settings for these?
--testp with --creative and just normal ice cream castle bubblegum stuff
So I'm going for concept art of a cyborg, and this is really beautiful except that the heart is a little out of place - but when I try to edit it, it says the image violates the content policies, no matter what prompt I submit (maybe because the image has nipples, but DALL-E put them in there?) I'm kind of afraid to keep trying, since I keep racking up violations
Honestly I'm kind of happy with this, but I do wish I could play with it some more
Yea unfortunately this is a known problem :/ but it's actually more towards the center of the body than most people think
haha thank you i will go with that 🙂
I wanted to experiment with this prompt too, here are a few if there are any you'd like
Lower left is really striking! Here's the other one I made that was a favorite
I do like the original a lot because it feels so near-sci-fi, the body is so familiar
definitely agree
Does anyone know if anything in the content policy would prohibit photoshopping my real face into an image generated by dall-e? e.g. I know we can't upload faces because there's no way to verify that we have the rights to use them, but can I manually add a face I do have the rights to use (i.e. my own) to a dall-e generated image?
(Obviously I would still be indicating that the image was AI-generated and all that)
You can do whatever you want with the generated image. I use mine as underpaintings all the time
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
does generating variations of an uploaded image not use credits?
no. it's a new diffusion process so it uses OpenAI resources, and takes a credit
Is there a way to add blank space around an image on mobile for Dalle to fill in?
Use the !pad command in #dall-e-bot
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
Oh perfect, thanks
Hey guys (:
I make music and I wanted to ask...
Can I use something I just did with DALL E as an artwork to my new track?
Means upload it to Spotify/Youtube etc
Yes.
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
Regarding that: It actually accepted it and gave me some clay-like renders of it back
whats these video dall e things ive seen on instagram? is there a way to play around with it
I can't seem to use outpainting, is it live yet?
Yeah same here
Outpainting is not directly a function but much more a
process that requires you to upload a padded image.
Have a look at the message below for more detail. 
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
I don't think that's what the announcement is saying. I've done manual inpainting a few months ago but this seems to promise a built in tool https://openai.com/blog/dall-e-introducing-outpainting/
We could do that before though no?
Look at #announcements
Just took a look at it, thanks! :)
Also worth noting the #dall-e-bot channel is locked for me 
It came completely out of the blue ^^
The #animation channel is a great place to find various types of animations.
There are currently a variety of tutorials available, so simply browse the channel
or ask one of our community's many helpful members in there! 
Does prompts in which name of well known real life person is mentioned gets flagged?
Yes, that's presumably the reason it got flagged, as you're currently not allowed to generate images with celebrities or real life people
If I try to get on this site it says it's an error https://labs.openai.com/editor
I know it's new, but I wanted to outpaint some things :/
So I literally just used my first 7 credits on DALL-E today to generate some images today and I got a batch of 3 per generation.
Literally 5 minutes later this was announced ( #announcements message ) stating now we have 4 per generation.
How can I reclaim my 4th pic per credit that I didn't had?
For the outpainting feature, is each square that is outpainted cost a credit?
For me as well, I get an "Page not found" error.
Yes
You cannot
Thanks
I solved my problem with GIMP (tool to edit pictures)
Since the site has the hug of death and I can't experiment; With outpainting can you change the prompt for each generation frame?
Does anyone know if they plan to reduce the credit price since they just reduced the text model prices? This new feature is great but can kinda cost a lot. I mean comparing it to a real artists making it, dalle 2 costs nothing. But it just seems high compared to the text AI.
Where are the big edits ? Are they stored somewhere ?
They are not stored, you will need to download them if you want to save them: https://help.openai.com/en/articles/6516417-dall-e-editor-guide
ok thx, i did that 🙂 But what if i want to continue to work on it ?
then you just reupload it on https://labs.openai.com/editor
ok got it, would be nice though if you can save the unused 3 others
Yes you can
how ?
I was replying to someone above, not saving the 3 others
ok 🙂
Not a queestion, but man what is this great! I was an artist 27 years ago and stopped when internet started. And now i am back in the arts, made some large prints. Probably an exhibition soon
Isnt technology wonderful? 🥰
is it possible to download all your created things from "My collection" without the watermark?
My friend claims he has seen this image of a toad before. That Dalle just put a filter on it. Is this true? Or is this literally a brand new image of a toad? Or is it a merge between existing toad images.
Is there any notation to force Dalle to focus on a specific portion or ensure that it draws a specific thing?
DALL E generates entirely new images using Artificial Intelligence, not filters
Check out #gpt-models, some of the guides there might help
He will not believe me. He claims Dalle's images are fake. I will let him be.
Dalle is for me, I guess. It's not for him. I believe Dalle because I've seen it for myself.
💖
is there any way to apply a filter to an uploaded photo?
how do I avoid getting distorted eyes and mouth in my photo portraits? I keep getting weird teeth and eye distortions
Hi! I'm not sure if I'm doing something wrong or if this is a bug but I can't seem to get the outpainting extention features to work. Whenever I extend the frame and try to generate the new parts it just adds a whole new image overtop of the whole thing. Any ideas on whats going on?
same here, would love to know how to fix this!!!
I may have figured it out. I was using Safari at first but then switched to Chrome and it worked. Seems to be a bug on their end with Safari.
yeah, just figured that out as well! cant wait to get to creating!
Yes it’s becoming apparent that safari may have an issue right now. Openai has been made aware of the problem. Please avoid safari for now.
Hi! Please tell me, where can I provide the feedback regarding the Dalle?
Thanks!
@long fjord @bitter isle see #announcements for an update on the Safari problem
Quick question, so I know what part of this to avoid. I was trying to generate an image using the prompt "cyberpunk bishop" and Dall-E reported that this may not follow content policy. Any insight as to what part of this is against policy? I read over it and think that the AI MIGHT assume it could be violent or political, but I'm not entirely sure.
does using outpainting consume credits?
I would also like to know
Hi! When I got my invitation to Dall-E, I accidentally tried to create an account with the wrong email address. As a result, my account is restricted, since my phone number is already in use with the wrong account. Please advise, thank you!
@short salmon does using outpainting consume credits? or is it free because it is just a test
It does cost credits
ok thx
@viral snow theres your answer
I have a question regarding outpainting
Do i get different results even if i use the same prompt, or does it always change regardless of what ur next prompts are?
Cuz i don't know if i need to keep on putting those words like "8k, highly detailed, photorealist, etc..." since Dall.E seems to know the context already, so like, if i want something different do i just need to describe it and add nothing else?
I don't understand this padding thing. Tells me to "Tap once with the edit brush anywhere in the padded area and then hit Generate" but the button is disabled. Meaning I still have to retype the prompt?
yes you need to retype the prompt, or put in a new one if you want the outside to be unique
thanks
Hello everyone, first thanks for sharing the art you generate, my question is the following, maybe it happens to you with the new functionality of Dalle(OUTPAINTING) that you press generate, the blue bar starts to move but out of nowhere it disappears and no generate any image?. Thank you
is there a way to publish things we've made on OpenAI? I know Night Cafe has a publishing feature and was wondering if this did too
how do you do outpainting? I am in editor with a piece that I'm trying to expand. Do I just ask it to generate what is outside of the selected portion?
Yes, but you must keep some part of the original image in the box so that the new generation has context of the style
is it possible to outpaint a head on this guy? the algorithm seems to hate the back of people's heads
the above comment made me think it could be possible?
do I move this slider? what does it do?
fail
I tried moving the slider up to the more zoomed out looking circle
Got the outpainting e-mail, tried it out, but it doesn't work. A rendering bar underneath the image gets somewhere between 1/4 to 1/2 of the way, then just stops and disappears. It doesn't use up a credit.
maybe post everything in bugs-and-issues
In your case, you can just Edit it, press F (for frame generation), select just enough or a bit more of the top portion
You can probably keep the same prompt you originally have too and it'll know how to draw a head
what do you mean press F? I tried that key outside of the prompt box and it does nothing
This.
Within the Edit menu, and this means the generation frame is active, you can drag that blue box around to "expand" your prompt
you may need to scroll down...?
there is no scroll bar?
might need to go to the website manually
https://labs.openai.com/editor
should unlock itself the next time
ok thanks. that looks different. do I then upload my image from the other editor?
no need, hopefully. Go back to your recents and find that photo you wanna edit
the new UI should be active now
Then yeah just expand beyond the photo but not so much as the whole blue square is outside, you're gonna end up with a 2-pic prompt collage
unless that's your thing
might be wrong channel but is there any way to upload images so they aren’t squared? like i want to upload an image with dimensions 849x1070 for the outpainting thing and want the whole rectangle. is this possible
I can't figure out how to get back 😦
Oh, you posted this twice. Here's my reply from the other channel: Yes, of course, you just need to upload your image from within the editor, not the "upload an image" link below the input field 🙂
back button is over there
I don't have that but I found it back in my browser history. however when I go back to the original image there aren't editor options or a scrollbar
I don't understand how they are separate, the url is https://labs.openai.com/ in both tabs I have open. I tried refreshing each tab. if I go to the editor from my recent list then I lose all my past art and there is no link back, I have to use browser back. I'm worried if I close my browser I will lose all my history

editor is an isolated window yes but idk why yours won't adapt the new UI in the default
best case is to just download and work from the editor separately
well where is the menu to view recent items? maybe I can start over. I'm stuck on the 1 painting I want to edit because I don't want to lose the page
by uploading
ok I tried uploading. and I added a frame, but it's stuck as a square frame so I can't stretch it taller than the image. how can I resize the frame?
this feature is definitely beta mode 😄 😄 even if I try to generate it I have to enter a prompt in this mode, it can't deal with being blank
what is the resolution upload limit for outpainting?
As of now, there is no limit resolution-wise.
So far, the largest image I've been able to upload was 20000x4500px, with a file size of 19.5mb.
However, there seems to be a file size restriction since I did not manage to upload an image of mine that was 50mb in size.
I hope this helps 
Hi, is there an option to change a uploaded picture to a different style? For example: change this photo to oil painting? I've read a guide, but I couldn't find the answer. Thank you!
Hiya, please keep in mind that the new Outpainting tool is only
intended to be used in combination with uploaded images.
To adjust the size of an uploaded image, drag one of the
corners to the inside or outside to decrease or increase it. 
(Image for reference)
Dall·E is currently unable to do style transfers.
If you'd like to propose such a feature, please
submit your idea(s) in the channel #suggestions :)
Hey guys I have a question: I tried using the regular editing feature without the intention to outpaint anything but it still generated 4 pictures instead of 3, will this feature remain or is it a bug?
How is this generated image getting blocked by the content restrictions?
it's probably detecting that it's semi-unclothed
Not sure how that's a problem? Unless it thinks the shorts are underwear?
Pfft, can't upload the generated image that I edited to get that one because discord thinks it's explicit because its less clothed than that one (middle of shorts missing due to ai glitch)
Hi guys! I am in my first month of using DALLE and wanted to know if the 50 free credits you start with roll over each month (they won't expire after my first month). I know that my 15 free prompts operate like this but wanted to ask someone who has had access for longer than a month 🙂
How do I use the new outpainting feature? Currently I was only able to use the standard edit feature, where you use a smaller image and write a new/extended prompt
Kindly refer to this message 
#openai-chatter message
Try accessing https://labs.openai.com/editor
Let me know if that works for you :)
is it possible to generate variations of uploaded images (without using the outpainting editor)?
Yes it is! 
Simply upload the desired image and click on Generate variations
(Have a look at the images below for more detail)
ah nice, thanks for the info!
oh actually wait is it possible to generate variations of any image or does it have to be a certain ratio
i keep uploading images but it tells me to crop them as if im going to put them into the outpainting editor straight away without giving the option of editing or generating
Select a square area and you will be able to edit or generate variations of your image afterwards :)
Generating variations is only possible for images in a 1:1 ratio as far as I know. That's why it requires you to select a square area.
oh ok
What do I have to press to use outpainting, instead of the normal editing? In the demo it seems that they create a generation frame and then click through three or so variations for that specific square (without spending credits for this specific sqaure) , but I'm currently only able to create the generation frame and write a prompt for that area and click generate, like before the update, where I get four new images, instead of one image with clickable options
is it possible to save an outpainted image to my collection?
not currently. the independent elements can be saved and will appear in your history, but you need to download the composite and store it yourself to preserve it
Hello everyone! I noticed that they added a feature called "outpainting" to expand on an existing image, but I've not been able to find where I can use it. Any help is much appreciated!
You can access the new feature via https://labs.openai.com/editor :)
If you visit https://labs.openai.com/editor you should be able to use the new outpainting feature.
(The image shows what the website should look like)
How does it work, can I say add onto existing image, or do I need to describe what I want added on?
You can add to an existing image by uploading it; however, you must describe what you wish to generate.
Have a look at this video to gain a broad idea of the new feature. 
https://twitter.com/_dschnurr/status/1565011278371794944
So, how come the prompt "Michael Jackson driving a little red corvette" is against the content policy? Is Michael Jackson considered sensitive?
Generating real people is against the content policy
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
I'm using the "editor" for the first time, and when I press "generate", it loads for a few seconds and then stops like nothing happened?
Try pressing the ? In the top right corner for a guide on how to use it. If that’s not the issue, come back and provide some more details and perhaps a screenshot
@short salmon haha, thanks. I didn't read far enough down the policy 'Do not attempt to create images of public figures (including celebrities).'
However, 'The artist Prince driving a little red corvette' does not give the warning, which is why I was surprised
Nothing’s perfect. It’s not like it’s a human doing checks on each prompt.
also: just because it's not blocked does not mean it's not against the content policy
hmm it doesn't detail why it would fail, I expected an error or reason (if I've run out of credits / it's against TOS for example), I'll try to figure out if it's one of those
@past geyser , which is a great point. And now I know.
ah yeah it's probably a TOS thing, even though it's not telling me that :)
It worked, thank you 😉
Thanks that's really cool, I think this is a great feature.
Outpainting generation is starting and not finishing. Servers overloaded? Tried two different browsers.
Hi all, can I make outpainting of a picture where there is a human face in it?
No, that would violate the content policy
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
do the 50 credits from the first month "rollover" to the next months ?
like if I have 50 credits this month would I have 65 or 15 credits next month ?
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
Is there a command to generate the default DALLE bot responses, by the way, or does it use text parsing to discern when it needs to be sent?
I saw that, but it's not clear if the first month's credits rollover
might be automated by a bot in the future
(I still don't know if they do)
@rugged wing Free credits do not roll over
oh
I've gotten errors on submitting two of my prompts that they didn't go through, so I tried again (as the error message says to do) but with the same result - however when I returned later, in my history multiple of the same prompt where there (and the credits were used, even though the error message said it wouldn't use any...)
Anywhere I can go to ask someone to fix this?
(lost 5 prompts this way so far)
Contact support@openai.com and explain what happened, they may be able to help you
alright, Thanks
thx
outpainting max size?
Do you mean in terms of images you can upload?
I mean max size after upload
I don't know. As far as I know there isn't any?
👍
leans over to peek at what @indigo prairie is up to
I just want to make one huge and detailed image
I'll go buy some credits then
❤️
Is it possible to change the email address associated with my account? I used my company email, but am switching jobs!
Contact support@openai.com
Has anyone figured out a way to use the outpainting feature that is not insanely credit-intensive?
i have a doubt , can i increase resolution in outpaintiing ?
i saw this as an example
This was done in via outpainting
Yes! Head on to DALL∙E's Edit feature! Read more here: https://openai.com/blog/dall-e-introducing-outpainting/ & https://help.openai.com/en/articles/6516417-dall-e-editor-guide
It's worth noting that that's like a 10-15 credit image, right?
Can I use the images created with DALL-E without necessarily identifying them or without copyright?
Can i publish my Dall-E 2 images on Instagram without getting in trouble with copyright or something?
do you know how to resize the mage ?
do you know how to resize the image ?
i got it
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
Yes.
probably by the time they finished it
Thanks!
anyone know how to remove watermarks in dalle image.
You can use an external image editor, or use the inspect feature of your browser to grab the image. It will be a webp image so you’ll want to use an image converter if you use the inspect feature of your browser.
You can also crop the image, which might be the easiest way, although you will lose part of the image.
can you explain the steps to remove water marks in image ?
what converter i need to use, how i can use ?
Sure!
These screenshots explain how to download the webp file
Outpainting Not working in my Mac I can upload image but then I can’t see new added images . On my iPad Pro I can go to edit upload but the tick for image won’t work
The converter I use is https://www.xnview.com/en/xnconvert/
This might help! If you're still having issues after looking through the guide, come back here
Hi thanks I know how to use but it’s not working on Mac and now iPad , hopefully it get sorted
It only works on Desktop right now, so it probably won't work on iPad
But it should work on you Mac
My Mac is midn2010 intel core using safari, but no it doesn’t work images missing etc
You can delete the watermark with an image editor, in Android you can do it with Snapseed by Google using the repair tool
how can i use outpainting?
Is there a way to separate character descriptions when there are 2 characters for the prompt?
I have a picture I produced with DALL-E and with external edit. I would like an higher resolution version of the same picture. Minimum variations are acceptable, but the more faithful the better. Is it possible to get it with DALL-E?
Can’t see image in generation frame
Yes and thank you
Why is DALLE 2 kinda, what's the word, mediocre when it comes to generating text?
there is a way to change my username on dalle2 profile? i would use my nickname instead of my real name
Most AI text to image models are right now, it will increase in accuracy as the models get larger :)
Please contact support@openai.com to ask about this issue
Lots of normal prompts like this seem to fail… can just do simple ones like ”cat”
Why is this?
Content policy enforcer seens broken…
From the content policy: Do not attempt to create images of public figures (including celebrities).
Why is that even against the rules?
For more info, please visit this link: https://labs.openai.com/policies/content-policy
Can someone tell me how outpainting works? Maybe it is the most stupid question today but when I place an additional frame to the image, what do I have to do next?
I can enter a prompt but it will generate an entirely new image next to my input image that is not an extension for sure
The frame needs to overlap part of the existing image to make a continuation of it
It only takes into account the part of the image that is inside of the frame, so the more it overlaps the more closely it will follow the style
Outpainting is the process of adding more content to an existing image with DALL∙E. Padded empty space is added around the image, and DALL∙E fills it in!
(There's much more you can do with outpainting, this is the most basic and commonly used. Check the #inpaint-outpaint channel for more ways to outpaint.)
Generate or select an image and use the !pad command in the #dall-e-bot channel.
Upload your image to DALL∙E and press 'Edit'. Do not forget to erase the watermark.
Generate it with a prompt that describes what you want to see in the new image as a whole.
okay that makes sense, thx
Hi, I am not sure about this one, so I want to ask to make sure. Are you allowed to use Images generated by Dall e 2 as album covers on streaming platforms like Spotify etc. ?
Yes.
Oh and is there a way to remove the watermark or do I have to keep it on the album cover?
objectively speaking, you could just remove it using Photoshop content aware fill f.e.
but not sure if that is intended
you can find the .wepb file which has no watermark: #community-help message
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
Take a look at the message above 
thx
thank you guys very much :) 
Hi, what credit package can I use to keep my generations private?
Yours are all private unless you choose to share them. This is not Midjourney. 🙂
That's so cool! awesome. Hope they could also work out the max size option of at least 5000px. Thanks!
hey everyone, what's your best recommendation for upscaling images to a printable resolution?
Well this is both...concerning and weird...
So.....what part of this is a content policy violation?
Maine coons? Or commodore uniforms?
the filter can be heavy handed sometimes - i once tried to make a painting of Fukuoka, Japan and received the same error
Hey, I got a lot of prompts and decided I wanna try outpainting, but for the life of me I can't figure it out
Have a look at this 
https://bit.ly/dictionairi-upscale
by @rugged wing
This should help 🙂
thanks
The word "coon" gets flagged since it's a racial slur.
Others had success with prompting it as just 'Maine cat' though, give that try 
i have a question about outpainting too - if i just want the image to be "extended" so to speak would I include most of the original image in frame and describe it to my best ability? is there any other method to do this?
Well that's unfortunate... But thanks, I'll give it a shot.
Hiya, just to clarify: Do you want to create an image that is larger than 1024x1024px,
or would you like to add more content to an image without altering its original size? :)
say I have a 400x600 image, I just want the ai to extend the scene to its best ability, kind of like the girl with pearl earring promo
I've fit it snug into the middle of my generation frame - admittedly I'm just low on credits and would like to make sure I don't waste one on getting it right
Then, to the best of your abilities, describe the image:
Include keywords that are relevant to the image as well!
If you're running low on credits, I can give it a try for you 
alright, thank you! and no worries, I can probably hold off for a month if need be.
@pine void Thanks again. Worked like a charm. :3
Sure thing! Also check out this tool:
https://colab.research.google.com/github/pharmapsychotic/clip-interrogator/blob/main/clip_interrogator.ipynb?authuser=1#scrollTo=30xPxDSDrJEl
It will generate keywords relevant to the image you have submitted, which may help you in getting a more accurate result :)
In outpainting, does pressing cancel give you back the credit?
It does not, once you hit the Generate button 1 credit will be removed from your balance. 
👍🏼
Does the outpainting feature costs my credits for anytime I expand the image?
Yep! It counts as one generation. 1 credit = 1 generation.
Thanks!
Hi there, I have a question. Is there any public API for Dalle2 available?
DALL·E is not yet available through an API, and users are permitted to access DALL·E exclusively via our public-facing website. We do not permit programmatic access via web scraping or data extraction, as outlined in our Terms of Use. We will continue to evaluate how we release this technology as we learn more during the beta.
@winter juniper
I would prefer the minimum size of 5000px. I don't mind if I need to pay more credits. There would be stages of processes such as, stage 1: composing in the standard resolution. Stage 2, generating variations or Outpainting, Stage 3, then when we're happy, we would pay additional credits for upscaling until we reach our desired image size. Just please don't make it that too expensive. 😁
Can someone explain to me the relation between different queries? How is a query influenced by previous queries? Only the previous or all the history? Does creating a new session relate to the process?
No relations at all between queries.
Any news on prices?
Thanks! I asked because many speak of "first try to get the subject, and later focus on the style" or "iteratively converge to a stable picture" and it was confusing me.
Is it possible to disconnect my Google account from Dall·E and use an email address to sign in?
Nothing as of now
Fair enough
If there is, you'll have to contact support@openai.com
Is there any way to save multiple versions of an outpainting?
I love them all and would love to have all of them in full resolution
Not exactly. but all the generations are in your history, so you can retrieve them there
Oh, will the version in my history contain the extended image?
As in, will it jsut be the new frame, or the entire thing?
In the last few weeks I have seen a lot of software being released for the creation of artwork. Has DALL-E already released the API or only for a private audience?
Don't most of these use Stable Diffusion?
[❓] — Q: Is DALL·E available through an API?
DALL·E is not yet available through an API, and users are permitted to access DALL·E exclusively via our public-facing website. We do not permit programmatic access via web scraping or data extraction, as outlined in our Terms of Use. We will continue to evaluate how we release this technology as we learn more during the beta.
I saw this official announcement but I saw at least 10-15 different apps like this one: https://www.jasper.ai/art
I don't believe that has any relation to DALL E or OpenAI
Got it! I found the results to be very similar.
from Jasper...
I have replicated on my account with the same description.
LIke I said, no relationship, as far as I know
Thank you m8 @short salmon
Alot of the edits I'm doing have text generated with them and I don't want that. Is there a way to keep this from happening with a command or something?
I think this issue has been tackled in #❌┃prompt-help before. Let me see if I can find it for you
Jasper claims to use dall-e, and uses a model trained on GPT-3
but given that api access to dall-e is prohibited i'm not sure how
Ok thank you! I'll scour that channel also
I only found a few things, mainly that its going to depend on the medium you're trying to generate, and that text should be easy to remove using the editor (aka inpainting), but of course that will require additional credits.
I did this yesterday, originally wanted poster art, but removed it to make it a full picture
I saw the same statement / news item and so I decided to ask here.
is it in any way possible to delete some entries from your recent session history?
or all you can do is make new session and have everything be deleted?
I believe your only option is to clear it entirely using the "New" button
Also for the future
We only allow discussions of other AI's in the #ai-discussions channel.
is there an estimated timeline as to when DE2 will be removed from beta and when (if at all) it will be open to the public?
is it possible to put my face in the ai and then tell put a cap on it or idk
I don't believe so, not at this time
Please review the content policy, uploading faces and generating faces of real people* is not allowed
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
nono i wann do my face in animation
Hey guys anyone know a place where I can find some instructions on how to use out painting? Didn’t fully understand the examples from IG
Anyway to make Dalle fit the images into 1:1 instead of cropping it?
Try Nvidia's stylegan3 for face related ai art
You can upload an image with just a border and leave it blank on the inside, and run your prompt on it.
why i didn't get month credits, or here is no month credits? my friend said its ~15 credits per month
I have a question, i must keep the watermark of the images to use it?
Have you gone below 15 in the past month?
nope
That's probably why
oh okey
Out of nowhere I got some free credits... Is this a bug? 😅 Was playing with dall-e. I had around 30, then, out of sudden, jumped to 74
https://openai.com/api/policies/sharing-publication/
i think this will help
lucky lol
Reload, probably browser error or something
If it's still there then free credits, if not you never had any to begin with
Nonono, I don't reload. There is hope... 😄
It won't go through when you reach 44
Did you report or request a refund at any time?
Nope, I did nothing
Huh
Well. I will not complaining... 😄
Ohhh okay ^^
It might be a bonus or smthin
I dint find anything about the watermark here...
this is what u searching for, i think hope at least
You get full commercial copyright with it which means you can remove the watermark and still get full access
Thank you guys
Know someone how many yousers DALL-2 have?
Yes. Avoiding deception is our primary intention. We're not enforcing that outputs are attributed to OpenAI / DALL•E 2. We'll update the content policy to make this clearer in the coming days!
You can read more about FAQs here.
Please refer to the message above.
Keep this in mind, especially if you are using your
Dall•E generation for a commercial project, since
it's technically still a grey area in the ToS. 
Please refer to the messages above too :)
I will
Just to be sure, I am wondering if I am missing something. I have found only two reasons to upload an image: to get variations of it without a prompt, or use a prompt to draw or redraw some parts, i.e. inpaint or outpaint. But is there some way to upload a picture and use the picture itself as part of a prompt, like "The subject of this picture, but with an anime style"?
What you're thinking of is known as 'Style Transfer'. DALL E does not currently have that as a feature.
Thanks!
eyo anyone have an idea on how it'll be with the credits when dall e goes public. ngl I kinda do like my free monthly credits lol
Why would this prompt trigger a violation of terms?
"In a packed stadium, two footballers from opposing teams, jump in mid air. They are trying to win an important headed challenge. The image is shot using a 1500mm lens with a shallow depth of field, f-stop f2. The crowd in the stadium are in a frenzy"
The only thing I can think of is footballers from opposing teams. Like I dunno, maybe somehow it considers that offensive. Everything else is basically just context
or maybe that crowd is in a frenzy, maybe it thought like violent frenzy
Could also be "shot" (which tends to make DALLE think of gunshots) or "headed".
I used "is shot" with some generations and it seemed fine
I'll try rephrasing it. Thanks for the input guys 👊✌️
It might depend, but I know for a fact I've had e.g. "a shot of a wolf in a forest" blocked while "a view of a wolf in a forest" was fine.
you can the image is taken with a 1500 lens etc maybe
Replaced "shot"with "image taken using...". Worked fine.
Thanks again guys.
Hey everyone, this might be an odd question, but I've been unable to generate a coffee cup made out of a potato, without there being a potato inside of the cup in all results, regardless of prompt wording, and I'm wondering if anyone else had similar issues ?
There's probably not enough images similar to your request in the training data to generate it
is there a way to resize an image, while using the edit feature, by typing in the desired size without having to adjust the size using a mouse ?
Where do I start with this stuff
Is there a database of all the images that DALLE has generated?
Not publicly accessible, no
There is not, you can only generate at the default size
Getting started guides are listed in #gpt-models
how can I check the generation historial without having credits?
your generations can be found on the right side of the website, or you can search your browser history
If you generate too much images does your history have a limit ?
cause i can't find my first try :/
I think it keeps the latest 50 images or so. I don't knoenthe exact amount 😦
Thx i'm gonna save everything from now on :/
it's a new month right? do i get 15 new credits
• You get 50 free credits your first month.
• 15 free credits will replenish every month after that, on the same day of the month.
• For example, if you signed up on August 3rd, your free credits will refill on September 3rd.
• If you joined on the 29th, 30th, or 31st of every month, your free credits will refill on the 28th of every month.
• Free credits don’t roll over to the next month, so they expire after month after they were granted – but you’ll get 15 new free credits.
• Paid credits will expire after 12 months from purchase.
so do any of my first 50 free credits roll over or no?
Free credits don’t roll over
i am based in colombia and dalle doesn't let me buy credits.. haven't been so sad in so long.
is there any way i can circumvent this? any reason why it doesn't let me buy them as an individual but it allows as a business?
Unsure, its most likely due to the payment processor's policies (Stripe).
im using outpainting on ipad and it refuses to work for me
how on earth do you use this thing none of the buttons are working
Outpainting is currently only supported on PC
Hey everyone! Haven't used Dall e in a couple weeks and wondering why they did away with the grid format when generating pictures. They used to be presented in a 2x2 square like this:
But recently they've started to be generated in a line like this:
Are we locked into that new format? Or is there a setting I can update to get the old format? Kinda sad cause the pictures were larger and clearer the old way, so I didn't have to download each image individually
That seems to be a bug, please transfer this to #bugs-and-issues, OpenAI staff actively monitor that channel.
Got it. Thanks!
Hi, I've seen images mentioning Matisse style paintings so I've tried to make Dall-E create an image in the style of Leonid Afremov. No matter how I try this, it gives me a warning?
Having AI generate images like this one would be incredible?😅
Best asked in #❌┃prompt-help - drop the prompt there and let’s see how to help 🙂
Also, I didnt receive a warning for leonid afremov testing here: https://labs.openai.com/s/KwSdRWfQ8kCQhiIgOm5RkBuw
Perhaps because you wrote "painted by" instead of "leonid afgemov style"🤔
maybe so! but it definitely gets his style!
It is more of an oil painting in comparison. It misses the bigger patches of oil where the light is "refracted"
Will paid credits which I got as a gift from OpenAI expired?
Since they are 'Paid Credits' they will expire after 12 months 
We paid the price of sweat and tears
Thanks for your answer
full body full shot photograph of Supergirl, female teen, cosplay
and
full body full shot photograph of Supergirl, teen girl, cosplay
gave me policy violations?
oh im stupid i forgot about full shot being banned
Q: can I use a generated image as my profile pic?
(asking because attribution would be difficult and I am wondering if that would be a problem)
edit:
nevermind, I think I figured it out with the bot channel...
Hi everyone, I was trying some outpainting and I'm getting policy violation strikes I can't quite understand. I'm uploading images that are my own and from my device with fairly straightforward prompts that seem to work fine in the "standard generation space"
It's not a massive issue but it did left me wondering, I would love to understand what's triggering them to avoid any future problems
For example, I am trying to make a contextual render for an object in one of the corridors of the International Space Station starting from a .png, and I keep getting striked.
That was a bit of a challenge as indeed there is a strong tendency for potatoes to appear in the image.
These were also not made of potato but rather ceramic. But at least they have no potato in the cup.
I'm sure this has been asked before, but I can't find the answer: is there any way to increase the size of the generation frame so that it can still capture adequate amounts of the source image when it's significantly further away?
1024px is pretty restrictive
There is not, currently
Thank you! How about for removing backgrounds? Is there an easier technique than "tap and pray?"
Could you elaborate? I’m not sure what you mean
Is there a prompt to have it try to insert certain text? say if I wanted it to say "Space" or even just and abbreviation like "SR" somewhere in the image
It seems like it tries but the letters get all jumbled and randomized with symbols
Hi, I have generated an image for a 2D character, and I want to animate that character for a Game, is there a way for generate the different views of an existing image, so that I can create a basic model to animate it? (PD: I'm new to modeling)
DALL E isn’t great with text. It’s technically possible but the more complex the word the more unlikely it is to get it right.
DALL E isn’t able to to generate the same character multiple times, unfortunately
Ok, thanks
maybe try "model sheet" or some other animation specific keyword with outcrop of your character on one side of the image
The only way to remove a background (even if it's a single color, like black) that I can find is to painstakingly draw (or tap, on a tablet) around your character portrait and very meticulously erase, then tell the AI what to do and hope it understands. Is there a simple way to get rid of the background in an image like this?
I want to have the AI put a different background in.
I would recommend downloading the image and using a 3rd party image editor select the background and remove it.
It would be MUCH easier than trying to manually erase the background.
Gotcha. And then, once re-uploaded, DALL-E can work with the transparent background?
Yep!
That's perfect. Thank you. Related question: this cartoony version of the character is a result of DALL-E generating this original image (she is not a real person; she was created entirely based on input strings) and then deciding she was "too realistic" and refusing to work with the image further. Shouldn't there be some way for DALL-E to recognize its own style and allowlist the images that it itself generated?
Hm. I’m not following again
So, when you try to upload this exact image on the web portal, you're met with an error message to the effect that pictures of real people aren't allowed to be uploaded.
However, DALL-E generated this image.
It's not a real person nor is it based on a real person.
Ahhhh. Yes, right now no faces are allowed to be uploaded. It’s a matter of privacy concerns and preventing deepfakes.
Yep. Unfortunately that’s just how it works.
Hmm. So that's a lot of credits I lost.
You shouldn’t have lost credits if it didn’t let you upload it
I lost a lot of credits creating the initial image before knowing that I couldn't re-upload it.
Ah. Well there is a workaround.
Perhaps DALL-E should be disallowed from generating images that it prohibits to be uploaded.
You can simply cut out or blur the face and upload it.
And then replace the face after
The face is unfortunately the most important part. 😜
This is a character portrait for a sci-fi RP.
It shouldn’t be difficult to cut out the face and put it back in later
I could help you with it if you’d like
Sure, that would be awesome. How would one go about doing so?
It’s just a matter of using an image editor to edit the face out
And thank you in advance for your time and consideration.
Right, I got that part, but how do you put the face back in if you can't upload it? DALL-E needs to be able to iterate on the whole thing, no?
I sent you a DM, it might be easier to collaborate on this there
OK - thanks!
I'm looking to use an image DALL-E generated for me as a wallpaper background on my pc, what would be the best/easiest way for me to make that image 1920x1080 without loosing quality?
thanks so much!
You're welcome! :)
@pliant wind see the link above that Smile just sent in that message
Handy, thanks
I heard some people mention that outpainting only produces 3 images. Does it produce 4 at its current state?
Right now, Dall•E will generate 4 images for you to choose from
when using the new Outpainting function 
you get 15 free credits a month, but they don't roll over from month to month
Hi, is there any way to give DALLE an image and have it generate a similar one?
Yea just gotta upload an image and generate variants
I'm working on a writing project and wanted to use some Dall-E artwork in it. Can I? And is there any specifics I need to keep in mind?
You can't deceive anyone into believing you created the image yourself. Aside from that, you should be fine. You should review the Content Policy as well as the Terms of Service too.
Content Policy: https://labs.openai.com/policies/content-policy
Terms of Service: https://labs.openai.com/policies/terms
API Terms & Policies: https://openai.com/api/policies/terms/
Thanks 

images are not generated on this server like midjourney ?
there's a website where you create and store images. Just look up Dall E 2 and you'll find it
is there a way to save more than 1 result from outpainting?
Nope, they are all generated on the actual website
They are all saved in your generation history, but you would have to edit it back into the image yourself
for the generation history, do you mean this on the right side? If so, I don't see them in there.
