#🗣|artisan-support-feedback
1 messages · Page 8 of 1
that one feels OK, maybe try it again? unless they are changing stuff that one should generate
looks like you had a space before /dream?
Yeah that message happens when it doesn't register as a slash command
I'll take a look to ensure nothing is wrong but might just be the underlying model you hit, styles appear to be generally working. Some models might veer into a certain style heavily though depending what I have up
oh just realized I broke it completely 😄
This gets asked a lot
I believe the most common answer is that no one knows and not anytime soon
No set date, not out of the question
do we train now sdxl turbo here or an other model ? @lethal haven
nevermind... i have seen the limitation of turbo and no,we dont train this "research" model here
when is controlnet coming back?
oh
also i noticed that the bots have trouble generating unordinary anatomy
like 6 legs
or multiple faces
hello I've been around asking stabledreamer for it's own opinion with this prompt :
hello stabledreamer please define or explain to me your own understanding of this word, {a keyword}
If I put different keywords regarding to subjects most of the result are horse, but if the keywords are object it's sometimes not horse.
any one have any idea why they (bots) love horse 🐴 so much?
You have the word "Stable" in your prompt? That could be why 🙂
ooooohhhh
what are the defaults for the bots in terms of CFG, steps, and sampler?
bot
that's private information and the bots are beta, they are always changing and tweaking stuff
How do you set the defualt to photo and not video. My prompts keeping entering as videos and is taking forever
Change format option to Image
Why some photos are blurry?
It may have been the 'front unzipped with plunging cleavage' that started it on a track to showing too much and its filters stopped it.
Why when I write wtf does the bot not understand that I need to draw?
Ok, that explains it, thanks for the message 
Is there an SDXL Turbo in the bot? Or when will it be included there?
turbo is not currently on the bots and we can only speculate if it will be : )
nsfw in pantheon: #🏅|pantheon message
Server down?
Hi everyone, <@&1106006823061966871> will be down for a bit for some maintenance. Back in a hour or so! 🤖 ⬇️
back yet?
It will be longer than expected, apologies
That's why I got msg of "application not respond"?
Indeed!
OK, thanks, have good day
please check the https://discord.com/channels/1002292111942635562/1047610792226340935
it will turn green when the bots are operational
Man this is wicked 😊👍
It seems there is a bug. I often get the message "Sorry, we can't regenerate this image at this time. Try prompting it again!" when I try to redo. I think it worked better before...
someone can fix this ?
I just had this 2 times
Watch the minutes without the image

Finally
This AI hate me

try it with fewer tokens to test, tends to not like all those long paragraphs
GPT?
3.4
Yes chef 🫡
I'll do that later
How to switch format?
is there a way to change clothes from an existing picture or change the face?
/dreamIllustration style, Hayao Miyazaki style, the theme is traveling around the world, three protagonists, healing, riding on an industrial style steam ball, flying very high, but you can see the scenery on the ground, soft colors, high quality, high resolution, 8
restrictions? where did you subscribe and to what services?
If your account balance is incorrectly showing 0, please wait 2-3 minutes to see if the system auto-corrects. If not, please confirm with your bank to ensure the transaction completed successfully & then fill out a request using the following form on DreamStudio so the team can investigate it further: https://beta.dreamstudio.ai/support
Since someone asked before - all images right now (not videos) on the bot are Turbo
How many steps does the Turbo model running on the bot use?
So according to this, it should be better quality than SDXL 1.0 50 steps?
So according to this, it should be better quality than SDXL 1.0 50 steps?
Who knows the solution to the problem: "NansException: A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check."?
There is 2min of cool down between each round, go around and vote during that period 🙂
why is there a random chance for a video? I specified the style as tile texture and it really said: lets make it ✨ change over time ✨
you can specify it also as video or image, look ate the other options
I may change it back to image by default today
Is it possible to create a PNG format ?
self reporting #1101178553900478464 message
i miss the controlnet so much
yes please
Is the bot using SDXL? It gives me much better completely different results than the supposed SDXL from playground.ai
the most recent info said sdxl turbo (#🗣|artisan-support-feedback message)
how to use seed? how to find the video by seed?
they are different models. playground's is "based on sdxl" but that could mean anything, finetunes, etc...
as stated a bit up the chat> #🗣|artisan-support-feedback message the bots are now running sdxl turbo, finetuned.
Good afternoon I have a question, is there some kind of language model in the bot? If it's not a secret, then what?
The bot prohibits generating an image if you don't vote. But how can I vote up if I don't like both images? Please give dislike so that we can continue to vote and generate..
Vote on other people's images.
But, I still want to be able to dislike something that is broken, distorted, or does not match the prompt
Also, I want to express my deep gratitude for this server. It's incredibly convenient, easy and understandable. And the images are very high quality! I'm amazed! Before this I used Mijorney. Yours is much better, and it’s free... I can’t stop being amazed. Thank you!
How do i generate? I type Prompt: but nothing happens just text
You need write /
dream prompt:
Change style, aspect, negative_prompt, format
profit 💪🤨
Hi, I am trying to dream. I tried in a few bot channels but doesnt work. It says dreaming but doesnt give the result.
I have the same problem 😦
Is this issue still happening? 🤔
I want to ask. Is there some kind of language model used in the bot?
I just noticed a reprompting lately
for text encoding? based on sdxl
No language model is in use at the current moment, but there could be one at any time.
It's happening sometimes. And an hour later or so it works again.
Gotcha thanks I'll look into it!
How to change the ratio of the robot channel
When will they connect? The result is simply stunning! Here is before and after the implementation of the language model:
The first is without a language model, and the second is with a language model, reprompting
there is no subscription for the discord bot
@lethal haven #🏅|pantheon message NSFW content and dv8shitpost voted for it
dv8shitpost is also in showdown right now with his own NSFW content
seems they know what they are doing when you look at the promts
are the bot images saved anywhere? or do they just disappear into the void from the channel when it is culled,
also, how is style enacted?
is it just an encoding used or
embedding(
Hi all! How to add image and reimage it with SD?
Bot images disappear after about 90 minutes. If you would like to save them, DM them to yourself or right click then select save image. The only images that stick around are those that make it to the pantheon.
Currently controlnet is not available. It may return in the future.
Prompts are collected at https://sddscrd.sfury.com/ but that's 1) very much wip 2) unofficial and 3) unreliable
I have a couple of questions. How many prompts are used? Why does he ignore some prompts every now and then?And the second question, is it possible to make a model so that when you click on Restyle, the style changes in order? Otherwise, now it’s random and it happens that it repeats the same style in a row when I press Restyle
#🏞|general-with-images message
Please help regarding this error
Thanks
Ton of boobs in the showdown
And I don’t mean artistic nudes I mean attempts at generating porn
They really need to use the AI, after delicate and intricate training, to go back and "detect" what it can, within images used for training. Not only looking for validation of subjects and phrases, but also self-rating and user-rating "best", "unique", "variant-of", "worst" and "invalid" for each image learned, and possibly words (as antonyms and synonyms) and phrases (word-pairs). Then, with the images evaluated, so most of the "worst" and "invalids" and the others... Build a new training set that focuses on the "uniques", and refines with the "variant-of" images... while using things like the "worst" and "invalids", as a form of filtering. (since the worst and invalids could be "other things", only the pointers to that data would need retention) Using hot-spots for odd "valid" images, where "part" of the subject is valid among invalid content.
So, something like "human heads" among "all heads"... Animals, fish, bugs, aliens, heads of needles, heads of genitals, head of lettuce, heads of robots, heads of class, heads of the bus... which may actually contain a "human head" somewhere in the image also... Are not influencing the creations, where they are not needed. Though, all are valid "heads", and may contain "humans", but not an actual "human head", or the "subject" wasn't "human heads", or even any heads, in essence.
These things also need "contextual interpretation", prior to "word association", for images. We need to stop trusting "limited and vague human descriptions", and develop formal, scientific, literal, and contextual pairings for what the contents of images are. Combined with hot-spots for specific content, one individual image could be use in training for MANY contextual phrases. Focusing more on the unique ones, and constraining the "variant-of" similar images as refinements, it would greatly improve all yields and even repeatability and predictability of creations.
Why is it Taking Too much time to generate an Image?
How do I generate drawings and not videos, please?
Change format option to image
please teach me how to generate good ai image generated like this
is it using lora or soething?
use good models
for example
also get detailed prompts
i'd suggest to ask ai to write sun for ya
Also you should know what style you want and what model would be appropriate for it
Like in this image, Cutecake A1 would work the best
@lethal haven See Bot 1
halo
whats up with the chronically bad humanoids in videos recently? so many double heads and chests
It'll be fixed up soon 🙂
Is it just me or are there way fewer bot posts per channel before they get removed? idk if the piles of failed commands and chats are clogging the channels? should they be reset/cleared?
I think everything is going fine? Depends on how many users and how split they are in the channels.
Video is going to keep undergoing some changes 🙃 Expect both better and worse outputs.
As a reminder to everyone, @hushed skiff is used to test experimental settings and models. It's here to help us further improve the models through various mechanisms of feedback. Sometimes you may see very low-quality outputs, but (most of the time) that's entirely intentional.
hi
I don't know who went and touch something they shouldn't smdh. Ugh
How to import LoRA in Stable diffusion or other models
This depends on where are you generating the images: Automatic1111 and so on... I'd suggest trying this at https://discord.com/channels/1002292111942635562/1002602742667280404
Not sure where to ask, but is the recording of the recent ama posted somewhere already?
Where control.net ☹️
Where ControlNET?
She was so good
How to make an object move in Video format? How to control this?
all videos don't work (
ahh wrong chat, anyone know whats wrong with this
like i didnt wanna make an ogre
LOL
I think you need to contact #🤝|tech-support
bet, thanks!!
Hello. Please tell me that I am not breaking the rules of the community by generating an image for the following request: "Orthodox church, church paraphernalia around, everything in gold, burning candles, dramatic atmosphere, semi-darkness, light only from candles, a girl of Slavic appearance is standing in the middle of the church, realistic portrait top quality, beautiful girl, white long hair, plump lips, wearing a white long dress, loose hair, straight, waist length portrait, oval face, green eyes, beautiful eyebrows, long eyelashes, light makeup, lips painted with red lipstick, realistic texture skin, realistic skin colors, photorealism, girl holding a Bible"
I don't see anything in that prompt that would break the rules. Sounds like you're describing a bride at a wedding.
thank you
*Man. It's Taking too much time to generate image
Please return ControlNET🙏. There's nothing else that's missing from him
What happened?
Image to text descrip????????
How to start
Go to one of the bot channels and use the / dream command (with no space)
Which model does the bot use?
is the stability text to image API down? I'm suddenly facing the following error
connection error: desc = "transport: Error while dialing: dial tcp: lookup stable-diffusion-v1-6.tenant-stabilityai-corweave.svc.tenant.chi.local on 10.135.192.10:53: no such host\
is anyone facing issues because of this?
where the hell do i start? i don't see any newbies servers
go to a bot channel, then type /, choose dream and enter your prompt. Also it is recommended you choose video or image from the format option.
any news for controlnet?
No news
A1111 🤤
Can anyone tell me which model and version is active on bots?
love from Chengdu
Thanks. can i ask which version? 1.0 or newer??
Newer is better
Oh yeah, thanks
You are welcome!Hope you have a good 2024!Love from Chengdu!
Can somebody tell me how the bots decide whether they turn my dream into a small clip or a still image?
it's randomized
Thanks!
So are there plans for SDXL 1.1 in the near term?
/drea
you can specify if you want:
how do I add a bot to my server?
Ah? Good to know, thanks!
Please tell me how to separate “green eyes” from “black and white tracksuit” so that there is no green in the suit, and the eyes are not discolored ?
That either needs better prompt understanding than SD(XL) currently has (best you can do in my experience is what you do already, specify colors for both), or better guidance (see https://rich-text-to-image.github.io/ for example, don't be fooled by fancy text colors, the solution is more generic), which probably isn't in the bot, though who knows, it might be as an experiment.
How to do image to image prompting with discord bot
Nobody responds to this question
@split finch @cerulean heath you cant, that because the bot that generates images,with the votes is used for training the new model version.
The training and votes are watched by the mods and admins to make sure no one would generate images against the rules AND to make sure many people vote to make it more democratic and not like an specific group of an specific server votes for specific images they want to train.
So the main reason is to control the training of the sdxl model in any way possible, to make sure no one is abusing it. you can do whatever you want offline in an tool the model is free to download.
for people who only want to generate without complex tool management i can tell you fooocus is very good and easy
It is not currently available. This is a test server and the devs are not testing that right now. They may do so again in the future.
It seems like the stable dreamer bot escaped its work cubicle. The authorities have been notified. I'll let you know when we have news about the situation.
EDIT : situation has been resolved, the chain is back on #1047610792226340935 message
Worst bug
How often has this been happening?
2 times, but after the 3rd time I got my image in DM
Will be generating again shortly
Did they remove the capability to upload image here like before?
The devs are not currently testing img-img/ controlnet. They may bring it back in the future. In the meantime, please help improve the bot by voting on images and videos.
hello how to define size ?
aspect:
aspect only define the shape
that's the only control you have over size
ok thanks
iinteresting
How do find other pantheon 24hr Leaderboards
how do we do img 2 img
hello how to define size ?
It annoys me that I try to tell bot to generate like a man pointing his thumbs down and he puts them up
Apparently the bot's can't draw a camel 😆
Models tab in SD "greyed" out. Any ideas why? Or how to fix.
图生图
laile
Does video work? None of the vids I got have any movement?
I have a problem, I can't generate an image anymore, it keeps saying "ERROR: @Wade already has a pending request:
Hang tight, you're still dreaming! :)" even though I haven't launched a prompt.
I have the same issue
Had it for a few hours now
Try choosing option format image
same issue here
Hey everyone, you should be good to go now
Please let me know if it happens again > 5min
There is an annoying bug lately:
Yes StableDreamer has breached containment once again. Authorities have been notified. I can't do any promise or deliver any ETA about its return.
@lethal haven Good afternoon. Why did the bot crash? Problems with servers? Or some other reason that should not appear ahead of time?
Like I said, we've pinged the geek squad already. Please refrain from pinging them about this otherwise they'll end up bombarded with pings. Thank you.
Should be back in action, sorry about that, unexpected server issue
Hi, I have a question. How can I do some image here? I try to do on StableDifussion website but is paid, can i do here free?
I get lots videos that have no movement, just a comment for the team. Also wow image gen only takes a few seconds, how did you do that??
^should we specify some movement like "zoom in"
feels like the bot is very biased towards adding women into every image
Nsfw in #1100484581037195384 message
Bot crashed?
bot down 
bot not working?
Bot down right?
@rocky vessel Bot is down.
bot not working?
/message bot not working
Does anyone know why the bot is not working?
Yup bot is asleep, gpu fixer squad has been notified. Please keep monitoring this channel and #1047610792226340935 for more informations.
Back up - will be intermittent as we perform an internal migration. Apologies for any downtime, the #1047610792226340935 will return to green when everything is finished.
cant generate, tells me I have still something in the pipeline
Hello I can't generate because I get the following message, it's been like this for hours: "Hang tight, you're still dreaming! :)"
Hi everyone! Currenty the bot is having intermittent issues because of an internal migration. Please be patient and always check the https://discord.com/channels/1002292111942635562/1047610792226340935
Regarding the #1047610792226340935 , if the icon is red 🔴, the bot is currently offline and not generating. Yellow 💛 means that the bot is intermittent and could fail sometimes, please be patient and re-run your prompts if needed. Green 💚 means the bot is 100% operational and fine.
Users that were stuck should be unstuck now
someone accidentally changed the order of these channels
Advice to Developer's of these bots from Me:
•You Can Remove Video Section from this Server as not Everyone Uses it.
•You Can Bring Back Controlnet to get More Traffic in This Server plus it helps people.
•You can Make New Server for Video And Can Try New Things to new server rather than This.
Make Sure to Provide Response Developers:
Thank you for you suggestions Mr. Fammu! I'm neither a developer nor with Stability but respectfully, this is a free-to-use beta bot, meaning it is a playground to test and develop new stuff that eventually becomes new models, workflows, etc. You have the privilege of using this amazing tool for free here while most competing products charge a fee or subscription. Seems like you want the server tailored to your needs when it's the other way around. If you don't like a feature, give feedback!, why don't you like it?, how do you think they can make it better? And if you like what you see, please vote it! helps.
I do Appreciate that this test Server is free of cost But as you saw many users need Controlnet to be Back.
yes, it's been requested many times and developers have said that it's not in the plans, maybe they are working on the next version! who knows, we just have to be patient and keep testing, generating cool stuff and voting.
🤖 <@&1106006823061966871> will be down for an extended period while we complete the migration as there will be too many periods of instability to keep it up during the process. Currently there is no definitive ETA. Will update when more information becomes available. 🤖 
Hi when should we expect the bot to be back?
There is no definitive ETA on the bots, migrations are never easy, everything will be up when it’s stable, please check the #1047610792226340935 and #🗣|artisan-support-feedback message
can't generate?, permission problems? please check both 🔴|bot-status and a few messages up in here 🙏
Further updates will be provided on Monday.
I have a problem too. My problem is that other people can't read.
what to do if i get- you do not have permission to send messages on this channel
read previous messages on this channel and check https://discord.com/channels/1002292111942635562/1047610792226340935
ok thank you
Well, if nothing else, I've come to learn how much I appreciate using your bots. Good luck over there.
Just wanted to say this server is awesome. I'll be glad when its back.
"When do you expect it to be available?"
Monday has already passed 🫠
Hey everyone, migration is complete but we're not quite ready to come back up yet. Making some additional changes. Hope to be back soon!
dang bro i gone for a while then this happened, i thought i got banned bruh..fyuh go go go team!! 💪
Where did it migrate?
South for the winter
May I ask what is the reason for the recent display that I do not have the permission to produce pictures, or is it temporarily unavailable to everyone
read posts above
read posts above.
read the pinned message above, temporarily unavailable
OK, thank you
same question
Hey, When will the Bots Arrive Again?
guess never
hi
Perhaps the testing is done?
hi
missing Stable Difusion
perhaps we get sdxl 2.0 now after nearly 1 year training per vote ???? i can dream
i hope, im craving a new SD version that understands prompts as good as dalle 3 does
that only needs an llm model attached to it and specific full text training
thats not so complicated
"anymore"
are the bots down? nvm i just read the messages
Come back as quickly as possible 🙏
Look at the message thats in every bot channel
yes i see, something wrong with the daw
hopefuly the controlnet will back tho
its almost been a year I wait for it
I see
🤞
hello, i have no permission to send messages in bot salons , can you help me ?
I mean how hard is it, all the Bot Channels literally have a message at the bottom saying the bot is down!
You expect people to read?
🔥
Is the bot down more than a week? 
Are you all stupid or just don't know how to read? It is written that the bot does not work.
why i cant prompt photos ? please help
bot.is.down
It is clear that it does not work, but others are interested in the question of the day when it will work.

but im up?
🍪
donnez des biscuits au pere noel pour un cadeau asap!! 
Yoo guys when will the bot be available it’s been a couple of days still not working?
dang i already missed this bot..come back..pls
the bot was down bruh
read bro
It will probably work when the new model arrives... On Thursday
i hope so..
I already thought that the bot was working, but it turns out it’s you)
haha
But we don't know how to read! 😭
That's why education is Important 😂😂😂
Rules for thee but not for me 😏
Everyone has the same problem, so I feel relieved
why 我 can‘t 发言
啊
Me too
hi !! , any news about bots ?
are all blocked?
I have the feeling they'll come back with a payment plan
maybe
Hey
Ugh...
the reason you had this sdxl generating bot for free was , it was trained on your feedback/votes, there will not be an payment for this because they need you for training the future models too. No one of you ever think about why the generations some times where so shit compared to sdxl generations in offline tools ? its because the parameters in the background where totally random. @uneven nest @raven comet
lol, this is how rumors start 😂
they will implement a new version of the sdxl model i think and we will generate and train it with votes again
its an fact
More likely poor devs just hit an unresolvable dependency chain, python is awsesome. But little update would be neat
oh i meant "it will come back paid" that's nonsense
@lethal haven can you pls explain to these "new" people here why we generate for free here and vote on the images ? thx
Or maybe the new teased model is so awesome they're trying to get it on the bot for testing, one can dream 🙂
the new model is as i know not based on SD
you mean what emad said
its not likely sdxl 2.0 its an diffrent sort of image generation
when we are unluck the model that emad teased is something that only runs on 16GB+ cards
The russian tweet about a new image model that does text so much better, and emads tweets like https://twitter.com/EMostaque/status/1753901219003707535
and what they use here is an version of sdxl
well, new model would still need testing right 🙂
not this way
as i know there are 4 approaches to image generation SD is only popular because its easy to set up and easy to train
but i'm still scarred by the last teased model that did text well :/
The future will be my witness.
#🗣|artisan-support-feedback message @remote siren @raven comet
there the proof
its annoying for me to search for things people are doubting
the training of the sdxl model trough human guidance was over 8 months long thats enough to get the same improvements as midjourneys has . Because most public available image generators with an voting system train their model this way
i think we will see a new sdxl version AND something that understand text like gtp3.5 and can generate images where text and details are better as sdxl BUT works totally diffrent so its not compatible with all the sd 1.5, 2.0 and sdxl models out there.
Only the future can tell.
are you ok ?
I'm fine and you ?
Guys, could I delete the models from the controlnet_sd15 version and just stick with the latest versions of the controlnet_v11? Or would the models stop working?
兄弟们,是不是永远不能用了?等了这么多天
I think you want #🤝|tech-support
凉凉了..彻底不能用了
不能FREE使用了?
I hope for the release of a new model by Thursday and the return of bot stability
请问公告在哪里看?
how to actvate stable diffusion
its offline for a migration
because I don't have permission for bots?
Hi world o/
Is that still what they say nowadays? .. in general .. it doesn’t matter
I have a question/in which section can I find out “why can’t I use the bot?” ?
這裡不能使用了嗎
Maybe you should ask for a refund
我也想问 这里怎么使用
inspect it will never come back
you have to make up your mind, it will never come back
凉凉了
??
这个有没有相关的教程推荐一下。。。。
Have the bot channels gone or is it just me/discord hiding stuff (this isn't a "why can't I use them" I know the bot is down)
yeah now theres only 5 bot channels
彻底不能用了。
They're not migrating, migration was completed 30th Jan according the bot status messages
"Hey everyone, migration is complete but we're not quite ready to come back up yet. Making some additional changes. Hope to be back soon!"
It's already two weeks
i miss you bot
if you love me too, you need to comeback asap
bet u dont
they hate all of us
this aint no stable since the controlnet is down
want the controlnet version back
when is the bot back? The bot chats do not appear to me
or they could've made the changes already, but they're waiting for a certain day before they get their new servers running
it was like that last year, when they pulled the sd 2.1 bot offline in january, only for it to come back in april to test out sdxl
We were warned back then that the testing was over, so that's different from now.
me too
的地方f
Как сделать картинкуу?
Пока никак. Бот отключен на неизвестный срок.
Are we hearing any announcements anywhere other than here?
/game_stats white background,1girl,pink hair,angel wings,white dress,glowing skin,arms up,
OK
By the way, StableCascade has already been released._.
Will February 24 really be the decisive day?
i dont think we will train SC here i think we will train SDXL 1.1 or 2.0 that can be used by normal pc users and not only GPU rent users
@noble abyss
Yoo guys anybody knows when will the bot be back to normal because I miss generating images
Less than 10 years
Still wait 2 weeks...
@lethal haven Hey buddy, Any update?
Hi, hoping to not sound like a complete noob, I have been trying to apply custom images on pose videos. Generated pose videos using Openpose/DWPose. Tried moore animate anyone. But always end up getting distorted faces. I do not want to use any UI (automatic 1111 or comfy ui). I want to try it out on python. Seeking help for this.
Emad mentioned February 20 and no further explanation. Maybe the bot will be working by Tuesday?
Or the new model been released?
No, several
#🤝|tech-support maybe. Really intended for the UIs, but maybe you can find a programmer there?
having the bot back up by the 20th would make a nice B-day present.
How have you been
How have you been?
I'm fine, thank you. And you?
the same, just awaiting for the bot to open!
aren't we all? 😁
Yup.
sheesh no signal for coming back? 😦
Hopefully I don't sound too sour, but you just can't say (paraphrased) "Bots go down for migration since keeping them on would mean too many interruptions" then a few days later say "migration was a success, but we want to make a few changes before the bots come back soon" and then..... nothing for weeks... The announcements don't really cover this prolonged outage. If there's been a change of plans, say so, hit a snag, say so, even just give a bullshit reason, at-least some sign of awareness people are kept waiting. My gripe really isn't that the bots are down, it's that there is insufficient communication about it.
I agree, they should be just open about it. This is silly.
Has the pantheon been migrated to Archived Communities?
If not, will it be?
ayooo, is dream studio down again????
waiting too
why i don't have bot
Because you haven't made bot
Probably never, all the old beta archives are also gone
I hope there are two or three days left...
2 WEEKS PROB
seems like it... super frustrating given the lack of communication around it
Why is the bot missing?
It's at school learning how to do fingers properly
don't forget that the bot is also learning how to write words properly and not some random bullshit gibberish
i cannot find bots to create my images...any help?
Why People's don't Read That Big Sign " BOT IS DOWN"
THE BIG YELLOW ONE’S THE SUN!
What sign?
I don't see a sign
😊
a little rabbit with green eyes eats a carrot::1 ice age::1 baby pink color:
Hello! Why am I missing bot 1,2,3? I can't access the image generation function
a
a little rabbit with green eyes eats a carrot::1 ice age::1 baby pink color:
Big rabbits are better
Hope
man i suffered so much without this bot
How did you make the pictures?
Why can't I enter commands to make pictures like before?
things have changed
1
@wintry siren cmon tell us something. whats going on ?
Dang i just read that too
I quicky login to discord to check this bot
And then..yeah disappointed again
-_-
Where are the bots???
Y'all are awesome and it's a privilege to be here.
🧐 Some of us are here to be jerks and make fun of others.
They have announced stable diffusion 3 , on Stability's official website, hopefully they will give us the opportunity to try it. They are very , but very interesting the quality of the images. It mentions that stable diffusion also serves as a base for Sora, the AI that generated video of amazing results that you have surely seen in the networks.
will it be free sd3??
yes, will be open source
and how this work? if sd 3 will release so in stable diffusion a1111 or comfy get in core sd3? or what
i signed up to join the waitlist
Do you guys think that if the bot comes to back will the prompt writing and selecting styles, aspect ratio & others will be the same to like earlier, but the only difference will be from SD2 updated to SD3? What you guys think about this ?
hmm
Can Anyone Recommend Me Best Free Text to Img AI? Except Bing
Dalle 3 Midjourney are the big two
They aren't Free
Nothing else, sorry mate. Stable Diffusion 3 is your only other alternative. Got to be on the waitlist for that one.
Oh, you said free. Then nothing except stable diffusion
Unfortunately I don't Have Graphic Card 🙂
So where is the bot ?
? @uneven nest we dont even know what we will train with the bot now, SD3 ? Lightning ?, SDXL 2 ? or SDC ? we dont know what they planned but we know votings and user gerations will also used for training in the future because MONEY. so ask the admins and devs.
Emad promised something would happen on Thursday
i guess i need to subscribe to midjourney
Still no bot update status since 02/06/2024. At this point is the bot going to come back online.
why is stable difussion 3 not here on the discord or am I missing something?
Pixart is a diffusion transformer, but be warned, due to safety they trained on a dataset with most faces blurred out
How do you use Stable Diffusion on Discord?
The BOT channel of painting is no longer existed. Which channel should you go to painting now?
nothing new under the sun
that's a masked link. don't click it
@cunning prawn @rocky vessel sorry for the ping, but could one of you remove the above message before someone clicks it?
No worries, thanks for the report 
what is masked link? It is IP grabber?
it's a link that's disguised as another link
.....................
长安书院
It's like when I post a Rick roll disguised as a Civitai URL, except usually it points to something you shouldn't download, not just a minor annoyance
[https⸵//www.civitai.com/models/907216](<https://m.youtube.com/watch?v=dQw4w9WgXcQ>)
goodbye everyone
time to switch to other alternative
cant wait for this one
discord member trained their bot trough voting images, but this is what we got
hi
Wow. I have long time not use that. And how to use it to make pics now
Hi. Server is under maintenance?
It was fun, while it lasted
Has stable diffusion gotten into some legal situation or are they about to sell it to another party
😦
i liked it the aesthetic of stable diffusion's AI, bing it's too artistic and midjourney its' not free
where are the good alternatives?
i want it
to late ?
/d ream
where is the command dream
怎么用的?
Sleeping
maybe if we all '/showlove '@ruby wing 🙏
SD3 bots are coming, why's everyone so impatient
will the discord bot will ever be online again ?
undoubtedly
[*] Stable BOT we will remember
you can use this site, base plan is free, it has many different stable diffusion models, including SDXL models. https://creator.nightcafe.studio/?ru=Arkhamcat
because theres been no update in the status channel since 6 Feb. loyal fans hoping for an update soon 🙏
soon as 6 march 2064
Hi #🗣|artisan-support-feedback, could you please consider adding auto-billing to Stablity's API? For example, if the credits drop below a certain threshold then a payment method gets automatically billed and credits replenish? That would be such a huge help.
well, loyal fans should know they're preparing sd3 for those on the waiting list, and in fact they've started invitations yesterday
how to can create image in discord
me too
This one looks closed
Bot doesn't work?
The bot works fine, it's just turned off 😋
RIP BOTS
Wow, still down after SD3 dropped.
Never again, my guess. They were here to improve an open source model.
Not sure why they kept the bot channels active after the release of SDXL.
Just be patient or use the alternatives https://www.reddit.com/r/StableDiffusion/comments/1ciyzn5/comment/l2ec9ji/
Yep, sorry about that, working on it!
Stability AI Membership - 1 Month Recurring - Professional Tier is completely different than that services?
I used discord account and landed Stable Assistant instead of Stable Artisan. I still unable to generate Stable Artisan cia artistan channel.
Correct that's different
Stable Assistant is used as the subscription backend for Artisan, as long as you signed up to assistant with a Discord account it should work on Artisan -- though we have some current bugs getting fixed as we go
if you did sign up with a Discord account and Artisan says you're not subscribed, you might just need to retry the bot again, OR you need to go to the Assistant page, hit Logout, then log back in with the Discord account, and then it'll work
yep already changed that
signed up for a $9 / mo plan with a free trial and got 50 credits on assistant instead of the 900 I expected
If you took the free trial, you got the 50 credits for now, and the 900 credits become available once you pay the $9 -- you can click "manage subscription" in the Stable Assistant page to upgrade from free to paid immediately at any time
We should make this more clear somehow
yep message will be more clear on that soon
thanks, that worked
consider using the tools available for more consistent output on SVD
will we be able to invite the bot to our own servers soon?
not right now but soon™️ yes (maybe a week or so)
It's too expensive. OK
This would work better if you could sign up with your email and then connect it to your discord account to use the bot.
I activated the plan on my existing SAI account (Google-based auth), but I see no option to link it to my Discord account - so dream command still doesn't work for me.
I want all these features in Stable Assistant too.
you have to do what I had to do. go to the site, log out, then when you go to log in again, use discord not google.
Prompt "portrait of a berserker girl, anime | oil on canvas" got Error: (ContentFiltered) - what the ....? guys at SAI, seriously - this is ridiculous, it's beyond complete bigotry level
I just signed up for the $9/month plan on the website, how do I link it to my discord?
/credits
/drean
can we generate in private or add a bot to our server?
I😒😒😒t does seem far too expensive, is the price likely to come down?
someone mentioned yesterday that this functionality might be a week or so out. i'm likely going to withhold any subscription until we're able to privately DM a bot rather than using the communal channels, so here's to hoping it comes sooner rather than later.
Sure it's been asked before but how does one utilize this to create video? The dream prompt only seems to be doing images.
you shouldn't post this in public - open a ticket in #1010934719455707218
@gilded spear are you around?
okay I deleted it
in all the channels you posted it in?
no two or three I will delete it all don't worry
reason why I left it in many channel was it's not clear where to leave it
but I had to let the people know, becuase many will have the same problem and it's related to money
not just for me
anyway I deleted my comments in two channels and will leave at mod-support
SD will habe to provide much better guide about subscribtion...it's very complicated
never post your personal information on a public channel - and open a ticket to customer support
Yooooooooooo need to fix filters a beautiful woman blocked simple a beautiful woman sitting at the ocean looking at her reflection
Filters will be getting an update.
I liked bots but I'm not interested in artisan. It's like midjourney but without the possibility to try first. So what are the alternatives?
You mean midjourney doesn't have the option to try first. This gives you three days. Still a no go at the price point for anything substantial.
it's still in testing. they're still training. They will still be releasing the weights. they already said why the price is what it is.
dang..after waitin 3 months, there will be no free generates..credits after credits
im broke
getting this again when I went to use it today. it was working the other day and I have a credit card entry to show I was billed.
when I go to login to the website to confirm that subscription, it keeps kicking me out.
log in to the site with discord not google
I did.
has me auth with discord creds, then gets to screen where i have to authorize discord allowing info to be shared with dreamstudio, then back to the page where it offers teh pricing plan.
it never keeps me logged in. clearly there's an issue with the federated login side of things, as it's not linking my sub here, and it's not letting me login anymore on the website. all of which worked 2 days ago.
Relayed internally to the people that handle that side of things
You don't, by chance, have something running that prevents cookies, do you? What browser are you on?
No browser cookies in discord. My discord account was associated with the paid account and apparently has become disassociated.
Browser cookies on the site
Reread
Ok my stuff is working again thanks.
I have. I carefully read your snark every post you make
question, how do I activate a private chat so I can generate images privately?
I think there's a genuine mixup there - you sent a message about SD3 weights release, in reply to a conversation in which andrei is comparing artisan to midjourney, and totes replied with a comment about midjourney note having a free trial equivalent to artisan's.
So I think you mixed up who you're replying to, or misunderstood the original conversation?
The option to use Artisan outside of this discord will be made available in about a week
i replied to this specific part of Totes comment - which was pointed directly at artisan
aah
Hello I activated a three day trial subscription but I don't know how to use that here on discord.
It's way too expensive right now
I generated a moderate amount of images over the past few days and burned through credits so quick
@split crown They made us wait for many months with no news in this discord, all to release a new paid bot. I find it dishonest not to have warned the users of what they were preparing.
Because of this practice alone, if I have to pay, I would go to another AI generator. (Unfortunately, my PC is not powerful enough to run a generator.)
IDK if this is an error of if the moderation is too strict, but I’m getting blurred results for SFW prompts — here, for example:
#artisan-1 message
How to upload an image and use control structure or control sketch, remove background by stable artisan?
ok I DELET, not gonna pay for bot steeling my job
they caught a cancer called capitalisme
I have the same question
Hi, I paid for the subscription, but I'm facing this message - You do not have a valid subscription. Please login with your Discord account while signing up to https://stability.ai/stable-artisan#choose-stable-artisan-plan.
Important: Make sure you click 'Continue with Discord' at the login screen! I attached my plan. Waiting for some instruction
Hi there! Is the email for your Discord account and Assistant/Stripe the same email? They must be the same at the moment to link successfully
will there be any free support for SD Artisan (maybe for previous versions like SD2) like there used to be ?
they are facing financial problems more severe than they had expected. there's a cash crunch... if they had cash, the weights would probably have been released already. don't think they planned on this any more than we did.
@shadow forge Thank you very much for your answer. I actually wasn't aware of this problem. As a result I searched the internet and came across this astonishing article https://elblog.pl/2024/04/03/stability-ai-faces-challenges-amidst-financial-struggles-and-leadership-change/ . I didn't think their financial problems could be this severe.
It's not the fact that the bot is now paid that bothers me since we all need an income to live and pay the bills, but the total lack of communication in this discord and on their site compared to their problems (I imagine, for fear of seeing even more investors run away). I was going to propose a donation system like Wikipedia which needs 25M dollars in donations/year. But the amounts presented in this article are indeed much larger and therefore more intimidating.
i do wish they communicated better, i agree
if it was launched with a "save open source AI image" angle with that message i think it would've been received quite differently
@shadow forge in relation to donations, I would also add that:
-
this AI generator is aimed at everyone, both the general public who comes there to have fun, as well as professionals who create and work on images for their business = very large audience reached throughout the world = a lack of potential donors. I myself wouldn't have minded giving for example 50 euros/year, if that could have allowed stable distribution to remain open and free for all, and I know that many people would have done the same here. I had also advertised (for free) to lots of people in my family and in other discords by showing my creations and I'm sure other people did the same when the bot was working.
-
By leaving the bot completely dead, and the site without news for months, there was a big leak of users, who went looking for AI elsewhere = big loss of potential customers and just as serious, big loss of trust ' some of the users.
I hope the stable broadcast will manage to recover from this.
Can't generate anything when I type in '/dream' please help
@drifting leaf bots are now paid on stable diffusion. I just discovered copilot which is free (requires a Microsoft account)
Ooooo thanks for info
I'm still mulling over this plan.
I was on a high-tier Midjourney plan for a year.
You get slightly over 600 images with the pro Stable plan if I'm not mistaken. That's not a bad deal, but it's not a great deal either.
With Midjourney standard you pay about $10 more than the pro Stable plan. But you get unlimited images when you switch to relax mode.
So it's a better deal, but it's nice to support a company than isn't doing so well by paying for Stable's plan.
Plus, I left Midjourney in the first place because I wanted to switch to generating Stable images on my GPU, but I wanted to do that with SD3 which I can't yet. At least Stable's plan would allow me to get a taste of it.
I just wish they were better at support. This place is like a ghost town. There are things I dislike about Midjourney but at least you're not on your own when something goes wrong.
But I think I'll still choose Stable in the end. They don't deserve the situation they're in. They've done far more for AI than Midjourney has and that means something to me.
They just need to start offering support roles to volunteers. I'm sure many people would offer their expertise.
hi, 2% and i are watching this channel for any technical issues
for billing issues atm we're just relaying that internally to the people that run that, we're looking into setting up a proper way to handle that
we do have volunteer helpers here, and they're pretty awesome
you'll find them with "community guide" or "community mod" role
for artisan specifically in here, unless i've missed something, we (2% and I) have handled every question that's come up thus far (not counting ones that somebody else got to first as eg with mas_flo above, or things unrelated eg sir_truc/clownshark discussing very-off-topic-to-artisan things above)
Ah, that's good to know, thank you.
@hollow shadow I completely understand your point of view. In my eyes, their big problem was a total lack of communication. As I wrote above, as a community, we should have helped them raise funds to relieve this company even if it meant contacting YouTubers to reach to as many people as possible.
Unfortunately months with no news about the not down = part of the community has been lost. And I completely agree with you, the unlimited offer of midjourney is very attractive and that of stable diffusion too limiting. It's also the unlimited offer that I was thinking of subscribing to for 2 months to test this AI. Unfortunately I find subscriptions very frustrating in general. I would much rather pay a large sum at once (150-250 euros or more) for lifetime enjoyment of a feature, rather than having to put extreme pressure on myself for 1 or 2 months to generate as many images as possible. for profitability of my payment. As a result, I don't dare take the bank card route because I say to myself "not this one, I won't have time to take full advantage of it, maybe in 3 months, etc."
This is why I would also like to create images on my PC (I saved up a long time ago to buy one). I wouldn't have time pressure or fear of wasting/running out of credits. But I don't know yet which PC to buy, it will also have to be able to run big fat gamer games.
I continue to discover copilot which is free and which gave me quite similar results to stable diffusion. I saw that there was censorship on the names of well-known fictional personalities and characters for understandable legal reasons (although it is frustrating). Oddly not all the characters are affected yet, and I was also able to put artist names in the prompts. One exemple for a same prompt, left = stable diffusion, right = copilot.
The one on the left looks like dr house in the Rick and Morty style
@winged mural yes true. These 2 characters share a deep inner distress which can be clearly seen in the image on the left. I really like making images of Rick in the style of Van Gogth, it goes very well as a style.
Hi, I'm playing with all the stability api functions, and wonder if there is a good workflow to turn a cartoon like human figure to a realistic human figure (keep the same pose). I tried inpaint & search&replace but not very sucessful.
Hi, I am wondering if there is way to use artisan in private dm, like Midjourney
not currently, for now it's locked to this discord - will be opened up to use elsewhere soon™️
how to prompt and where?
Hello! how to change aspect ratio? default 1:1
in the #1237459938901491852 channels.
click the +5 more link after you put in a prompt to get a menu, one of the options is aspect ratio
⭐️⭐️⭐️
Stable Artisan is producing server errors. Is there any timeframe on fixing it?
looks like that was an upstream issue with API that was resolved before I even saw that question lol
Anyone here training custom LLMs that wants to chat? Im working on several projects and need to talk to other people in this field
Do the api and the discord artisan need 2 seperate subscriptions? The api says I'm out of credits, yet I can still create images here in discord (and there's no way I rean out after only 4 images).
tag, you're it 😉
yeah api and artisan are separate
/crea
/settings
How do you buy more credits?
dear SAI, could you PLEASE move bigots far, far away from content moderation system?
that's not bigotry that's just a bug, it's a known bug and being worked on
@soft sluice here's one for the books. i created this image with the text to image tool on the stable video website, then tried to upload it to the image to video tool - and the site refused to animate the image it just made, telling me it was unsafe content.
How do you buy more credits for artisan???
Hello everyone
if you login to the page where you subscribed I believe from there you can upgrade to a higher subscription tier
er -- specifically go to https://assistant.stability.ai/ login with your discord, click "manage subscription", and you can "update plan"
Thanks I swear that wasn't working last week
Hey, just a heads up - Stable Assistant seems to be experiencing an outage, requests are failing with "I apologize, I've encountered an internal error. Please retry your request, and I will do my best to fulfill it. If the issue persists, please let me know so I can address it accordingly."
🎨 @dense bluff generated 2 images with SD3, using prompt Editorial fashion photography in high resolution: a beautiful female model with dark skin and intricate tattoos on her face and neck, aggressive pose, wearing a metallic, metallic high heels, and a wide-brimmed black hat, against a harsh desert landscape with mountains in the background , high contrast lighting creating a dramatic, tense mood, forward facing pose, shot on Ilford Delta 3200, monochrome black and white palette, seed 1451366694, aspect 9:16! 🎨
I think Assistant has a support form somewhere, the Assistant team isn't on discord
Thanks, it seems to work fine again
My new credits does not show in the discord interface
Do I have to use the artisan channels? Can I not privately DM the bot? I don’t want my images to be public.
Artisan will be available to use outside of this discord ... at some point, but not yet
i previously said "soon" about that topic but priorities have gotten distracted & focused on SD3 launch recently
Ahh rip. Would’ve happily paid for use if it wasn’t in a public server but I guess I’ll wait till weights are released and I can access it where I access everything anyway 😅
im trying to sign up to use the Stable Artisan discord bot
i tried logging in with discord but it said 'account verification required' even though my discord account is verified
then i tried logging in with google, it worked, and i put in my credit card. but it brought me to stable assistant chat bot
and when i try using one of the artisan channels on the SD server it says i don't have permission
what do i do?
I had this problem. I logged into stabily.ai with my discord account.
生成一个带有科技感的产品树
hi everyone
?
Hello everyone, Can I ask what's the difference between standard diffusion and Midjourney? Which one is better for new users?
Generations with Artisan now use Ultra by default!
Ultra is like Core, but including the latest-n-greatest experimental archs -- so right now it's using SD3-8B in an advanced workflow
You can use regular SD3 or Core at any time in the /dream command by just using the Model argument
Midjourney is in another castle
This is stable diffusion. use /dream to gen in one of the #1237460191780540456 channels. you can edit things afterwards and stuff. have fun.
assistant and artisan share credits. if you successfully got credits on assistant (check your profile in top right) those credits are being used on artisan
SD3 needs work with fingers and toes (the correct number of) 😄
Also needs work with tail placement for furry character turnaround sheets!
Also, ultra wide ar doesn't seem to work with ultra
Hi, im using Stabilty API and was wondering if it’s possible to get email when we use for example 90% of the credits?
stablediffusion is out since 3 years now, don't ask, they won't be there
midjourney is 10000000000000000 x Better than stablebullshiter

comfyAnnonomous just added SD 3 support to ComfyUI yesterday on github and pushed the changes
Thank you, my friend
1boy,(Cute,handsome,wearing outdoor sportswear :0.7), 3D,(Face close-up :1.2), (at night, surrounded by glowing plants, flowers, flying fireflies, bonfires), (Ultra detailed, aesthetic, beautiful composition, rich bright colors, volumetric soft light).
Have you tried SD3 yet?
is there a bot to add to my server?
they haven't said they've released that yet, just that there will be one
cheers
are u dumb >?
midjourney has answered this multiple times.
you have to generate in the #1237459938901491852 channels
Please make support for Stable Artisan to be universally used in Stable Assistant, there is a Javascript discord error some of us receive when having to log in the current way that prevents us from accessing Stable Artisan on Discord because it continually does a failed login loop. I've tried everything, and now I will have to drop my subscription because I can't access Stable Artisan specifically on Discord.
is the artisan bot still coming to other servers? i wanna sub but i dont wanna use the bot here let me take it home'
Yes, we're working on making it available on other servers. Apologies for the wait.
Apologize for the trouble, and thank you for letting us know. If you don't mind, could you provide a screenshot of the error here or via DM? We'd like to try to reproduce the error and investigate further.
^ hiroki here is the new project manager for Artisan, he's getting things organized so we can launch Artisan properly!
welcome to the team
I signed up for stable artisan via the website but it says I dont have access when I try to execute a generation... only seem to have access to stable assistant. Is there a setting to link it? Seems like a very bad UX design if there were extra steps that I now cannot access.
Sorry for the inconvenience. Could you please check if the email address used for your subscription is the same as the one linked to Discord account? If you still can't generate images even though the accounts are the same, providing more details (e.g. the specific error message) would be helpful for further investigation. (We do recognize that the login process should be improved...
yes, its the same email...
"You do not have a valid subscription. Please login with your Discord account while signing up to https://stability.ai/stable-artisan#choose-stable-artisan-plan.
Important: Make sure you click 'Continue with Discord' at the login screen!"
All well and good to tell me that now,except I've signed up, and I specifically signed up for artisan.
Thanks for checking! Just in case, did you sign in with Discord (not Google) to subscribe? If you logged in with Discord for the subscription and used the same e-mail address as your Discord account, it should be available. Please let us investigate further in that case...
I clicked sign up for artisan, and followed the prompts. there was never any documentation that asked me to sign in to discord. VERY bad UX. The first I see of this command was attempting to use what I just apparently signed up for.
Thank you for the feedback. We definitely need to improve the login process.
ok but can we please link my account so I can use what I signed up for?
If you can log in to Stable Assistant, could you please try to link the account with following steps?
- Log in to Stable Assistant https://assistant.stability.ai/
- Log out from Stable Assistan from the menu at top-right icon
- Go back to Stable Artisan page https://stability.ai/stable-artisan, then login with Discord account which is linked to the same e-mail address as the subscription
If this doesn't solve the problem or the case the e-mail address of the subscription and Discord are different, further operation may be required. In that case, sorry to bother you, but could please you send a support ticket from here?
https://stableassistant.freshdesk.com/support/home
Thank you for your cooperation.
how do I cancel my plan?
Hi, to cancel your subscription, please visit the Stable Assistant page, as your subscription is shared with Stable Assistant.
Login to your account on the website (https://assistant.stability.ai/login), click the top right button that shows the first letter of your email address. Then click “Manage Subscription,” and on the top right, click “Cancel plan.” Confirm your cancellation by clicking “Cancel plan.”
Just for asking, we Will be able to use the finetuned models of SD3 (2B, 4B, 8B) of the community with artisan?
1
these are the availble models with artisan
/dr
{'Code': 2001, 'Msg': {'error_code': '58000', 'error_msg': 'INVALID_CLIENT_IP', 'data': {'client_ip': '3.0.169.119'}}, 'Data': {'error_code': '58000', 'error_msg': 'INVALID_CLIENT_IP', 'data': {'client_ip': '3.0.169.119'}}}"
Why did this error occur
Could you share what prompt and parameter you tried that made this error?
Hi, Thank you for always using Stable Artisan!
I have just released an improvement of the Upscale in Stable Artisan. With this update, when you set the Creativity level to 50 or below for Upscale, the conservative upscale method will be used, allowing you to upscale images without altering their content. Conversely, by setting the Creativity level to 51 or higher, you can continue to use the creative upscale method as before. I hope this update will be helpful to you!
Are Ultra and SD3 8B the same thing?
no
Ultra is a workflow based on SD3 8B
they are not identical
Ok. Well, hopefully we get access to 8B some day and even that special workflow to run locally on our PCs.
👌👍👊
Thank you for your comment. I will do my best to make Artisan better and fun!
Hi guys, sorry for asking this question here. But I wonder if your midjourney bot is working? Mine shows "scheduled maintenance"
cuz I didn't join midjourney's server, I wonder how long whould the maintenance take
me too, maybe 1 week.
mine is working now!
Can someone help me? I can't log in to the website and this thing keeps trying to charge me $118
which website?
can anyone tell me how to do this?
https://www.youtube.com/watch?v=kw1GP_wGsAk&t=319s
did you follow his steps?
how to ask for an image?
Hi, you can use /dream command and input prompt to generate your favorite image.
hes not making a tutorial yet
Dress Up
https://youtube.com/shorts/L_FxtWsuM2k?si=G5z65n94hMIlwUl5 Please share your thoughts. I would appreciate it from ppl who know the technologies. XD
URLError: <urlopen error [Errno 11004] getaddrinfo failed>
I'm wondering which Models (SD 1.x compatible hopefully) are best for generating UI UX designs for websites
Anyone knows if there is a way to control the video generation? Or does it generate itself randomly?
Can you invite the stable artisan bot to your own channel?
Hi, basically Ultra is the best.
Hi, currently the image-to-video feature does not support control. However, we are continuously working on improvements.
Currently, no, but we're working on making it available on other servers. It's taking longer than expected, but we are making progress. Apologies for the wait...
Thank you for your reply.
On a related note I guess, given how Midjourney is testing an interface on a webpage, which may appear to be applicable to a greater audience with ease of use, is this something being planned too by Stability AI - which may seem to suggest a shift of focus?
ultra? Which model is that?
Thank you for your reply! We believe that ease of use is very important in addition to providing the technology itself, and we have several plans in the works. Stay tuned and look forward to them!
you can choose the model with model parameter, but it will use Ultra as default.
I'm trying to use it locally (since its free).. do I need the SD3-medium model then?
Yes, Ultra is only available via the API or Stable Artisan/Assistant. But you need SD3-medium to run locally.
Hi, I created those two images and wanted to get them merged.
Target is to keep the three girls and the space ship.
How should I do?
create
can you give me a link
Hi, I created those two images and wanted to get them merged.
Target is to keep the three girls and the space ship.
How should I do?
how can i get the free ?
You guys use stripe, right? If so, is it possible I can arrange a date that I can contact you to verify my identity to get my account back because? turns out Google Workspace does not handle Migration of email that well and I really run from a business email after. moving from email to email. simply just clicking delete on my old domain. Unfortunately, yeah, I cannot access anything that email receives now and I'm locked out and cannot cancel the subscription.
I'm afraid you're confused. This is artisan.
huggingface is what ur looking for
and their SD3 post
You can start with 3 days free trial by choosing subscription.
https://stability.ai/stable-artisan
Can you send me the email, please?
Dming you a list, as I would not know which one it would be not being able to check no more. One moment please
(5 at most, google auto logins got me)
I need to know the one you have a stripe sub for that your want cancelled.
messaged
but I'll post the domains,
sudo-rm-rf.dev
alinuxa.com
Would be what it would be under, I sent you the two, I suspect were what I signed up with
I'll post it publicly, for convienence sakes, as if anyone wants to bother scouting out my emails, scraping. more than happy to hand out consultations for IT
Are you still working on this? Because I'm on the phone with Google one right now
I'm just on hold, but they told me that they can actually possibly forward me to the workspace teams and have it restored. I just don't want to waste any of your time So I'm handing you a quick little update here that I'm likely able to get access back from the sounds of it.
Thank you for using comcom analytics.
"comcom analytics" supports all community managers (moderators and server owners) by stats, visualization, and analytics.
If you have any questions, feel free to ask us!
Your dashboard
Help
Support server
Other languages
en: help
ja: help Japanese
Hi, thank you for using Stable Artisan!
We're excited to announce that we've added a variety of new features.
Previously, image generation was only supported through the /dream slash command. However, we've now added the ability to upload and edit images using the following slash commands:
/search_replace
/remove_background
/erase
/inpainting
/upscale
/outpainting
/control_structure
/control_sketch
You can also view a list of all available slash commands by using /available_commands.
We hope you enjoy these new features and continue to use Stable Artisan!
After signing up and logging in with my gmail account, how to I link my login to my discord account?
Nevermind, problem was "artisan-1" was not enabled at all, but artisan 2+ were for prompting..
Nope, nevermind, still can't access the bots because I did not sign up with my discord account...how infuriating.
Hi, sorry for inconvenience.
Have you signed up for our subscription with 3 days free trial?
https://stability.ai/stable-artisan#choose-stable-artisan-plan
Stable Artisan is a fun multimodal generative AI Discord bot that utilizes the products on the Stability AI Platform API within the Discord ecosystem.
I have signed up for the 3 days free trial, but when I signed up I did so unknowling with gmail account thinking I could link my discord account easily. I then say the warnings of needing to signup with Discord account on this discord channel.
hello
Where is the bot ?
read #artisan-faq
Hi, thank you for always using Stable Artisan!
We have added a new command, /style, to Stable Artisan. By executing this command and specifying a style image and prompt, you can generate an image that matches the specified style. You can create images like a Van Gogh-style cat or a Japanese-style spaceship.
Please enjoy!
Thank you for trying the /style command. Since my previous description of the command was not clear, let me share details.
By using the /style command, you can generate an image that mimics the style of a specified input image. Please attach an image of your preferred style. Describe the content of the image you want to create in the prompt.
The /style command also has an optional parameter called fidelity. Fidelity can take a value between 1 and 100, with a default value of 50. Setting a higher fidelity will result in the style of the input image being more strongly reflected. Conversely, setting a lower fidelity will result in a weaker style influence.
For reference, the input image for the style, the prompt, and the fidelity setting can be specified as shown in the attached screenshot. In the screenshot, the image style is Gogh's The Starry Night, the image to generate is "cute cat" in the prompt and the fidelity is set to 80. So the command should result in a cute cat image with quite strong style of The Starry Night.
/style
@sly vault can't get anything to generate in the artisan channels today
We are sorry for trouble. The issue should have been fixed and you can generate images.
thank you, much appreciated
Hi,
Thank you for using Stable Artisan!
We have updated the reroll button for the /dream command to launch a modal that allows you to edit various parameters (such as the prompt, number of images, model, etc.) set with /dream. With this update, you can now adjust specific parameters when rerolling the /dream command.
We hope you continue to enjoy using Stable Artisan.
Hello
I have a question please
I am trying to use Stable Difussion from this server
[2:42 PM]
and this is that I get You do not have a valid subscription. Please login with your Discord account while signing up to https://stability.ai/stable-artisan#choose-stable-artisan-plan.
Important: Make sure you click 'Continue with Discord' at the login screen!
Hi, sorry for the inconvenience.
Have you signed up for our subscription from https://stability.ai/stable-artisan#choose-stable-artisan-plan.
If you haven't signed up, please register using the same Discord account that you use for the Artisan bot.
If you have signed up with a Google account, please follow the steps below to link your Discord account:
- Log in to Stable Assistant https://assistant.stability.ai/
- Log out from Stable Assistant using the menu icon at the top-right
- Go back to Stable Artisan page https://stability.ai/stable-artisan, then log in with Discord account which is linked to the same e-mail address as the subscription
Stable Artisan is a fun multimodal generative AI Discord bot that utilizes the products on the Stability AI Platform API within the Discord ecosystem.
Sorry for the late response. If you have signed up with Discord account which has the same e-mail address as your Discord account to use Stable Artisan, and you still have the same error message, we need further investigation of your account status.
In that case, could you please send a support ticket from here?
https://stableassistant.freshdesk.com/support/home
How to I give a prompt?
Thank you for using Stable Artisan.
You can generate an image with slash command /dream and add prompt parameter.
You can also do /help and /faq command to see a quick help.
Also /available_commands can be used to see available slash commands.
/available_commands
Is Stable Difussion only available through paid plans now?
You can use image generation based on Stable Diffusion model through Stable Artisan, and you have 3 day free trial for new user.
https://stability.ai/stable-artisan#choose-stable-artisan-plan
You can also get Stable Diffusion models from HuggingFace under a license defined for each model.
https://huggingface.co/collections/stabilityai/image-6669941c4646eeac61f1704b
Stable Artisan is a fun multimodal generative AI Discord bot that utilizes the products on the Stability AI Platform API within the Discord ecosystem.
how can i cancel my subscription?email support doesnt answerand you keep trying to charge my card every few daysim not able to cancel it from website because, it keeps redirecting me to provide new payment detailsurl in email about unsuccessfull payment:
https://assistant.stability.ai/
i click, log in and it redirect me to:
https://stability.ai/stable-assistant#choose-stable-assistant-plan
Sorry for the inconvienicne. We'll check the status with support team.
@here
Thank you for always using Stable Artisan.
We are pleased to announce the release of search and recolor on Stable Artisan.
With search and recolor, you can select a part of the original image using a search prompt and repaint it in the color specified by the prompt.
You can use the search and recolor from /search_recolor and Search and Recolor button.
In the example, I took a Japanese painting depicting waves and Mount Fuji as a base image, and recolored the waves in purple and yellow.
- search_prompt: large wave
- prompt: the large wave is colored gradation of (bright purple: 1.0) and (yellow green: 1.0)
Additionally, I changed the background sky to shocking pink.
- search_prompt: Background sky
- prompt: The background sky is colored (shocking pink: 1.0)
We hope you will continue to enjoy using Stable Artisan.
Nice
was this meant to be in the support-feedback channel?
haven't used it but ok.
What is going on here
brutal
Who?
NOBODY ASKED

I have no idea what stable artisan is
thanks bro

Photoshop can do that with a sliders and a color picker. Come on, now.
stability ai is still around?
admin used the wrong channel for announcements

What was this about "Free and open software" again...?
RELEASE A GOOD STABLE DIFFUSION MODEL ALREADY
Is this something to do with Flux?? Then we don't really care tbh...
I thought this was a better version of SD3
I always use stable artisian 
Stable artisian stable artisian. Name is in brain now. Yes advertise to me please
no, more than a color picker, at least
Dont need it, we have flux
wake me up when a good model comes out
fuck artisan, who’s actually paying??? nobody
flux >
have a nice evening
Thank you for always using Stable Artisan.
Bruh, I haven't even heard of that garbage until now 💀
what's flux
I'm hoping flux gets automatic1111 support soon. I don't like using comfy.
me too ngl
I still need to download flux
cosXL edit
Use forge
use swarmUI!!!!!!
Sameee, the inpainting tool makes it all worth it
ComfyUI the goat
darn
sup bud
remember when we could, I don’t know, use the free channels
I wish there were improvements made to allow modern models to run on AMD.
Ok yall are hating on it now just because they pinged. It looks perfectly fine 😭
Both comfyui and Automatic1111 have pros and cons
Who tf reacted with the nose and Israeli flag, that's CRAZY
We don't have pro, which is significantly better than dev
yeah it's fine but nobody is talking about it
This chat was dead 2 seconds ago 
I've thought about forge and swarmui....but I have everything set up in auto. I'll maybe wait a few weeks/months and if it never gets support I'll switch to forge of swarm.
yes
ok this was fun bye guys
what is flux?
I can simply not do what I do in ComfyUI in automatic1111 in any capacity.
please just use swarmui!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Insane compression artifacts and blurry shit. It does not look fine. Fix your eyes
But several times better than SD3
The best new open source model that has ever been relased. Amazing quality. It understands what a human looks like.
Isn't this just linear color transfer and segmentation? Like the code I wrote here over 5 years ago: https://github.com/ProGamerGov/Neural-Tools
Anything is better than SD3 💀
stable-diffusion.cpp
when was it released lmao
The best model atm
17/10/24
Like last month
Couple weeks ago? Maybe 1-2 weeks ago.
(X)
I use that 111 thing, Idk about stable-diffusion.c++
Fix your screen maybe it looks fine on mine? I aint zooming in for every detail!
It's not. The "best" is pro version, which is not open source. Dev lags behind
Looks mid, try better next time 👍
Lol it’s horrible
why did the japanese characters in the third pic change completely
oh ok
i really fell out of ai image news
ewwwwwww
Nah ideogram is best if were talking closed source models
Then you're blind as fuck. I have a good screen, if anyone needs to fix their screen, it's probably you who's looking at a tiny ass phone who can't see anything
lol
bingus
Yeah that's why I said "open source". It would be cool if they release the pro version....but even if they don't, flux is so much better than anything else right now. Hopefully people can finetune it and create loras for it. Then it will really get cooking.
Oh okay thank you for informing me
Lol why even say this? What's the purpose?
the pic went from 1080p to 720p to 480p, in that order from pic 1->3
Because it’s not good
To correct misinformation. It does not look good in any way.
Are you 12?
are you?


Not really, SDXL is better than Dev
discord is savage
Sounds like you are
For prompt adherance? Not that I've seen.....
i'll wait for proper finetunes, from text, prompt adherence etc it seems superior but it lacks stylism sdxl finetunes have
Sure in that case, but quality is the most important shit. No use in having it follow your prompt if the output looks like shit
Fair point lol
i got @ed overwhere. but cant find it.