#💬│general
1 messages · Page 503 of 1
Yes
good morning fellow bauble hunters!
i'll be back in a half hour
@green crystal feed so ruined, I had no other choice.
That bauble too quick to me
Bauble
No they dont chat
Bauble
So I'm guessing the Perplexity was updated to no longer pin your model. If you want to use something specific you now have to manually select it. This is disappointing, it's not that I don't like Sonar it's great for Deep Research and Labs, some of my spaces hav tools that really benefit from pinning Grok.
Boom
I like the manual selection option, but it's just me
that's nice when you're not copyi pasting into a data data tool that I need to be able to use quickly without extra clicks
for web searches from comet it's one thing but not remembering it for other chats is another
5 min cd + 40 messages. That was what we were seeing yesterday. Although one rare one dropped outside the cadence.

We has 6 monutes on 75 messages before
Its inconsistent today
Baubleee now
bauble
Now
The message count didn't start until the cd was done. There's been 15 messages or emojis so far.
But the cd is random
in 3
ooo mad kitty!
Was bugging so much
wassup mate
I'm guessing Perplexity's GPT-5.2 bill got too high lol
Everyone wanted to try out the new model lul
Yo is there a Perplexity bot on Discord
Nope, but you can make your own: https://docs.perplexity.ai/cookbook/examples/discord-py-bot/README
I feel like I have a Chance to be in the top 10 bauble
Baubleee
Same
wtf
??
Same
Failed
gah, hate when that happens
No one collected this item
@green crystal at this point I'm starting to think that they should get rid of the non reasoning version. https://fxtwitter.com/DaveShapi/status/2002443575380824260
Damn, doing economic and historical research, GPT 5.2 is markedly dumber than even Gemini Fast...
︀︀
︀︀OpenAI has completely lost the plot, screwed the pooch, and whatever other platitude you want to use.
Are the models intentionally nerfed after launch window passes?
Might as well use gpt 5 nano high instead of non reasoning 5.2
This has happened several times. Broken on their side.
We need a make-up bauble!!!!
Who to ping?
Yeah non reasoning is terrible lol
Ooof
And they're using that for the smear campaign.
Use a chain of thought prompt
Guys I have the pro version of perplexity how do I get the role?
Use invite via browser.
no one
Click the pinned chat
It will be hilarious if they keep doing it when oai has their ipo
Retail would get wrecked because they'd try to short it just to end up as food for institutions
Wait for the next one I guess
I don't see it :/
@green crystal free tier chatgpt research is dog shit. https://chatgpt.com/share/694804f2-7c14-8013-ad0d-990ab65797bf
It can't reason enough lol
yo chatgpt is great at making game cheats
Click #💬│general go to the pin tab
This is worse than grok btw.
i just learnt to reverse an .apk i just added my mods in mini militia
Baubleee
Bauble
Grok actually did decent.
Grok might achieve AGI in the next 5 years
Finally in the top ten, just have to keep holding on until 25th
Does mine shows up?
Yes it does
OpenAI made deals to purchase ~40% of the global raw, undiced DRAM wafer output until 2029.
Bauble
#Bauble?
bauble squared
In 3
Why can u gibe people pro for free?
?
Now
:\
Bauble
Like a free Pro subscription

Not Fate
@viral quest u there?
Hmmm
Camper
My lucky charm
I cut my finger on tongs. typing has to be speech to text for a bit until the bleeding stops. 
😭

Get well soon.
thanks it's just a little cut but it's annoying because it's on my right hand.
Yeah. Understandable.
Hi
Do Perplexity have option of languages like Tulu Kodava Toda Konkani Chitpavani Beary etc?
Someone correct me if I'm wrong, but I think as long as the underlying provider's model supports it, then the responses will come back that way, too... even if the translations in the user interface aren't completely there.
So the app should still be usable if any of the underlying models support those languages, even if Perplexity itself doesn't support it in terms of UI translation for the interface to use it. Does that make sense?
Yup
There isnt enough messages here
Nothing much
Same same
Been bed rotting
Pretty much
@solemn folio hell bro
I love what u doing
Bauble
Could u give me more details in dm‘s?



Lucky charm
Lucky ducky
Do eeet
Now
Santa is home
Shall the bauble come
Please don't spam guys, we have spam protection 
Is this considered spamming?
I tought spamming is like a letter over and over
Oceanic wsp bro
It seems like it can auto-mute for long replies sometimes too... but the mods are relatively quick to recover when it's an accident. It's happened to me when I wrote stuff up a couple of times.
I see
So if you find yourself timed out after a long response... be patient and mods should unmute you.
How so fast
It spawned with his name on it already
Mod ninja
@halcyon coral please mute @halcyon coral
s!mute icelavaman 1h spamming
Where can we suggest stuff?
stuff for?
Perplexity

\o/ a stay of execution!
a holiday pardon
Is it somehow possible to get unlimited images generations?
on perplexity or you mean generally?
Perplexits is great for images
I love it
Nope,
I hate that
Is there any ai that makes unlimited pictures as good as perplexity?
I dont mind paying a bit
Free users getting ~2-3 images/day, Pro users aiming for ~100/day (often fewer during high demand, ~20-50 reported), and Ultra users up to 1000/day, but expect dynamic, frequently changing caps due to high demand, with free tiers getting basic access and paid users hitting temporary throttles, defaulting to older models when limits are reached.
oh
Boss i had 2 pics today till limit said stop i am pr user
just use the underlying image models
fal.ai for api stuff
freepik
midjourney
from gemini for nano banana + veo
from openai for their image models + sora
The limit in Gemini right now works in mysterious way
I dont understand shi, i‘m new to intern ai shi
zimage is super fast and qwen should be coming out with new edit model next week if you have a good gpu and can run locally
or rent cloud 5090 for like 40 cents an hour etc
I‘m phone
ah then deffo cloud stuffs
https://apps.apple.com/app/id6752555050
is this the real banana nano?
No
When did Perplexity go full "Black Mirror S07E01 - Common People" on the subscription model?
Can u dm me with the things u said and explain if u want to do it ofc i dont want to annoy
I‘l take a look
There's some things to try... quality will probably just not be quite as good as premium models. There's all kinds of models that run on many consumer graphics cards on places like huggingface and civit. You seemed to wanted to be doing edits, which only recently became something that models can do and retain original consistency without too many changes otherwise. So if you're looking for quality and consistency I can only think of a few models that are ranked highly to try locally... Z-Image, Qwen Image and Flux (technically a suite of models). Z-Image generates images similar to Nano Banana's quality, with a focus on professional photography (so I don't know about styles). Another that gives pretty good consistent edits is Qwen Image, according to what I've seen. Flux has been out awhile and is known for being more realistic. However, both Z-Image and Qwen Image requires about 16 GB VRAM to use (I can't run them). Flux can theoretically run in less - around 12GB, but you'll still want 16 GB for the higher quality versions of it. With this, you can get at least relatively close. Z-Image and Qwen can surpass Nano for specific use cases sometimes, but not all the time.
If you're serious about doing unlimited images and want something in line with what you're seeing that is even close to Nano Banana or ChatGPT Image, and can't pay for third party services which claim unlimited image generations like Higgsfield or an alternative, you'll need to save up to invest in a decent graphics card with at least 16 GB VRAM. Generations will still be slower than premium services, though... but at that point it's just the electricity cost and maintenance/normal upkeep/wear-and-tear.
So my 4090 with 24GB can do the job?
I am on phone😔
I really appreciate u message
❤️
getting a used 4090/24GB might be an option for some.
Definitely should
you can also just rent cloud gpus
runpod, saladcloud vast lambdalabs etc
much better electricity cost wise depending on where you live
and if you're not sure you will continue to gen a lot each day etc
That's a very solid recommendation. Would also answer the questions around whether it's doing what you want it for, and give you some time to work with it to figure out how to get better outputs.
Bauble
yeah if the cost of the gpu is 2k for example
most people's usage won't actually recoup that in the 2 years before they'd want to upgrade to the next gen of gpu
so depending on usage and other hardware they have, cloud is better
buying ram these days is insane
Gib prompt
(And model selected)
Give
Ohhh u german
Nope.
I send this to perplex and sayid recreate with bmw
Bauble?
Bauble
Now
hello
Baubleee timeeee
Skill issue. Mine nailed it. @naive sage
hallo
I love getting new bauble variants. Gotta catch 'em all.
But that was your prompt.
Ah, so you wanted more like - replace this car with a BMW (prompting is very picky/precise)
As in you want a bmw car?
Yeah
What bmw car did perplexity make then?
What model is this?
No model he mixed the mercedes with the bmw
But ngl the bmw looks hella good like that
what image gen model is this
Idk perplexity
Not nanobanana
what do you have selected
Nanobana made this
Bro
Russian songs are as follows:
I have to teach nano right now
You gotta select the image model in perplexity.
Uhh i chose automtic
Apparently you don't know how to select image model in perplexity.
Uhh nope
I think thats the best nano can do
Wait. Nanobanana pro is max exclusive?
ye
Understandable, lol.
Yeah it was pretty pricey to use their API when I checked.
its pretty expensive on the api
Whats api?
I‘m new to this whole stuff
😭
its what perplexity uses to access googles image generation model
How can i get it?
nano banana pro?
It's too expensive. Don't bother.
Pay as you go type shit.
@wet lintel learn about tokens first. Then look up token prices for ai models online.
Token? Like memecoin?
No. More like credit points to measure ai usage.
Also, it's very hard to regulate usage.
So you might end up burning 10 dollars in a single prompt if you use something like gpt 5 pro on api.
No. Go to open router.
What it this?

Thats slicl
anyone fro isreal
The button went off right in front of my click
thx
Not an AI-related question, also a very nuanced one that goes beyond one word answers.
Get max
If your priorities are image generation, go for Gemini ultra
Its bad
Look at this shi
Baubles are appearing fast, can't keep up
sorry
yup they come faster when there's natural convos here
Standard practice
It's capable of generating images like you want, but definitely have to spend time learning prompting. In this initial phase of getting started with image generation, it's probably not to best to immediately go all out and get all the tooling (my brother gets this kind of mindset, and then never uses it). I'd advise subscribing to all the free services you can for image gen on all the main providers, you'll get a lot of daily credits between them. The models you can select from are pretty much the same from other services, so you can experiment how best to prompt different models, read articles on how to get higher quality results, how to make more consistent edits, and so on... then, when you've gotten your feet wet, and a feel for your realistic use case (what you're trying to do, how many generations that might take monthly, etc)... then you'll be in a better state to grow from there.
For example, look at the attached prompt, it's from a video walkthrough of how to prompt GPT Image... specifying things like camera model and imperfections greatly enhances some model outputs (Gemini does good with preserving imperfections and making things look more realistic, but GPT Image tends to need that if you're going for lifelike images).
I would almost guarantee that Nano Banana (especially Pro) can definitely do what you're wanting to do here... it just takes some time to learn the tools. When prompting straight from Gemini, if you're using 'Fast' mode, it's using the base Nano Banana model. If you use 'Pro' it uses Nano Banana Pro. (Not sure what it uses when you use 'Thinking')
Should have asked before subscribing, after is like hey just trying to collect info
@naive sage Suppose I am using Gemini AI Pro, which includes a standard image generation limit of 15 images per day. If I do not use those 15 generations on a given day, do they roll over and increase my limit on the following day?
I don't think they rollover
Not to my knowledge, but Gemini AI Pro should have a daily limit of more than 15:
I certainly don't get 100 images per day it's more like 15
You might be using it in AI Studio, maybe - which is different. The free use ight be around 15 there, not sure... but otherwise it requires an API key and isn't tied to the AI Pro subscription.
I know they say it's subject to throttling depending on demand, but oof... that's a lot lower.
And the limit resets in 10 hrs
🤏 😎
Based on the current service status for Nano Banana Pro (powered by the Gemini 3 Pro Image model) as of December 2025, here are your specific limits:
Daily Limit: Nano Banana Pro
- Official Quota: 100 images per day
- Note: This applies to the standard Google AI Pro ($19.99/mo) subscription.
- Current "Real-World" Cap: ~20–50 images per day
- Warning: Due to high server congestion in December 2025, many Pro users are reporting getting throttled significantly earlier than the official 100-image limit. If you hit this "soft cap," the system may downgrade you to the standard model or block generation until the reset.
- Reset Time: Rolling 24-hour window (often resets around 2 AM local server time for many regions).
Comparison by Tier
If you are on a different plan, your limits are:
| Plan | Daily Image Limit | Resolution |
|---|---|---|
| Free Tier | 2–3 images (Strictly limited due to high demand) | ~1 MP |
| Pro Plan | ~100 images (Subject to congestion) | Up to 4K |
| Ultra Plan | ~1,000 images | Up to 4K (Priority Access) |
Technical Specs (Pro Tier) - Resolution: Up to 4K (2048x2048 or similar aspect ratios).
- Reference Inputs: You can upload up to 14 reference images at once for style/character consistency.
- Characters: Can maintain consistency for up to 5 distinct people in a single generation.
User Note: If you are consistently hitting the limit early (e.g., after 20 images), try switching your "Model" setting from "Thinking" (which uses more compute/Pro quota) to "Fast" (standard Nano Banana) for draft work, saving your Pro quota for final renders.
Would you like me to help you optimize a prompt to get the best result within your limited daily quota?
Use share text io
My bad, next time
What is the context window of gemini 3 pro in perplexity enterprise pro plan?
Oh, actually there is no memory features for enterprise pro plan, so struggling
What video generation model dies perplexity use by default
veo 3.1 fast
Enterprise Pro is no different from normal Pro, what you get extra is Data Privacy & control and Organization/Team feature
Pro can only generate 2k images with NBP
Ahh i see
I can only generate 3 videos a month tho
Sucks
Video gen is very very expensive
Oh my god not sure how to tell you that, memories context is not there is enterprise pro plans, it can't even refer old conversations context.
Gemini gives 3 per day
it's actually different
Google can eat the cost
I can confirm that as an enterprise user
#💬│general message and that's not even counting image gen and daily llm use
They would go bankrupt
Not an enterprise user myself, but from what I know, data is erased after 7 days so memory feature is not possible
Bauuuble
Yeah, they are definitely eating some cost on that (if people are using it every day)... their API for 3.1+Audio costs: 40c/second. Basically 2 bucks for even a short 5 second clip.
🙌
Hi Kari, any other options do i have? Chatgpt teams plan I'm fed up, gemini still sucks in answering from google drive search, perplexity is excellent but there is no memory feature for enterprise plan
Perplexity works on galaxy note 2
that's cost for customers, not cost for google themselves.
you can configure automatic file deletion, threads are not deleted
I don't think memory data gets erased after 7 days
try gemini business
Currently I'm having it, it really bad in searching data on Google drive
You can probably still use it on the web browser
It should, I guess.
On galaxy s5?
It's too old
web browser
are you deciding to buy an old Samsung based on whether it can run Perplexity App or not?
Why are you asking this? I'm pretty certain it has to be android 10 and above
I have too much old phones
Im talking from a s9
the new billing info is sorta poop
It's usually Android 10 and above
Samsung experience 9 and Android 11?
Are u guys using these AI for business or just fun and trying experiments?
damn
Mostly personal usage, but it does go beyond fun.. research/personal projects too. Sometimes work questions, though work doesn't technically allow us to use AI yet for development (we are so behind the times) but as long as I don't paste in code it's OK.
Get yourself 128GB DDR5 ram for sweet sweet price of $1500 and save yourself some trouble
by running models locally
like devstral 2
Get 5 mac minis and run Kimi K2 locally
@naive sage why no one has any business idea's to utilise these AI models to make money, I mean everyone just want to try it to solve their homeworks, projects, maths, Instagram post, edit photos, but no one is coming up with the idea of development or to start a business using these models. What the hell
that's way too expensive
and not effective given input cost
It depends on what you need to do
Perplexity "is" one of the business idea to use AI models and make money
tho they are in loss
There are lots of people who use them for that, but the way AI usually recommends to do it is usually scummy. You still need to come up with a novel idea which uses it and integrates it well, otherwise it's just another frontend to existing services at a markup with no value (I suspect that's why there's hundreds of those AI reseller sites that look like knockoffs. Many are even technically illegal b/c they squat on trademarked domain name/service names)
well there's Andon Labs who is experiementing with the LLM running a vending machine, it hasn't gone so well
using AI for personal productivity is literally making money tho?
it's doing your job faster, easier, and better
@naive sage more than these AI models u get lot of help from youtubers, these AI are ok to create or generate videos, photos of imagination which a graphic designer can do, atleast a youtuber tells u to go here n there ull make money, but AI is just repeating the Youtubers language in text format. AI doesn't have their own ideas like ironman's assistant Xavier, i dnt know how to explain.
youtubers are just hype-slop
Can anyone explain to me that what is Ollama?
some are in it for the fun and education
but it's mostly just copy pasting trends and slinging slop
these generalized LLMs are no good, but specialized domain-specific small language models would be great to be used in business! This eliminates the huge costs associated with big general models, 99% of model always goes waste.
Can anyone explain to me that what is Ollama?
Application to run models locally on your pc
ollama is a way to run models locally and plug into that instead of openai api
using LLaMa.cpp
Does it need alot of ram
Yes sir
depends on the model
you need atleast 64GB of ram to run anything decent
What about Claude for coding on my computer
you need atleast 8GB vram gpu to get decent speeds
It define what type of " money " you're talking about here, for me what i get from AI is knowledge, time and the ability to learn more and faster than i normally do without AI, that's money right there, it provides me a great foundation and support for me so i can focus on what generate real value better when the time goes on which is knowledge
Nothing you run locally will ever come close to Claude
What’s vram
Best I could think of is Qwen 3 coder
Time = Money, AI saves us some time! 😌
Gemini 3 Pro shines here
So the way how i see AI, you don't need to use it only for making for " money " so that it will be useful or valuable
the gap for open models is 6-7 months behind closed source
so for specialized domains, you definitely can lora or finetune a model which does better than claude
Just use it on what u need, things that will support u later on
If you need Claude, get a Claude subscription, no model that you run locally will ever come close
Video RAM, basically Graphics Card with atleast 8GB memory
do you think AI is applied a lot in cybersecurity?
because I might use it for a university project fuck it
It's def used in cybersecurity
Like I am looking for a AI to help me code
Hello there
fuck it, I’ll try creating something original
Get a Windsurf/Cursor subscription
something worth working on for 4-6 years
gl
чё вы
Please use English, thanks
glm is also usable with claude code for much cheaper
thank you, this shit will definitely be important in 6 years from now
I don’t got money for AI
Check out Gemini CLI, they offer free quota at the expense of your data
Yes, you would need a gpu with 4gb of vram and 32gb of ram along with strong cpu which in this case is my i7 11800h on my laptop
Take a look at this: https://arxiv.org/abs/2512.02654 - Ars recently had an article about AI's being integral to cybersecurity as well. People use it to complement traditional detection systems, find patterns/suspicious entries in large amounts of logs and stuff like that... lots of applications for it.
Are Capture-the-Flag competitions obsolete? In 2025, Cybersecurity AI (CAI) systematically conquered some of the world's most prestigious hacking competitions, achieving Rank #1 at multiple events and consistently outperforming thousands of human teams. Across five major circuits-HTB's AI vs Humans, Cyber Apocalypse (8,129 teams), Dragos OT CTF,...
Running all of them with 100%
U can use gpt-oss-20b
+1, GLM is best Open replacement of Claude
with 4-5 tokens per sec
- Kimi K2
this model is really bad
kimi is great too ya
Ik but it's fun to test out
gpt-oss has got terrible alignment
+1
I always used Kimi for fun roleplay & writing 🤭
you should run the 120b version
It's great for everything!
thinking* variant
idk...
Isn't the 5060 pretty good (I know nothing about PCs)
i mean it's pretty decent
you just need to double the ram
then 120b will run at 0.5 t/s
Hahah
There are many free option for u
aren't the same guy who spent $200 for a subscription? 💀
If u willing to search
it's
nearly
as my pc
bro

the ram inflation is crazy
hehe
what is this
erm it's 128gb
I am able to run Devstral 2 using 3060ti + 5070ti (completely loaded into VRAM with almost 32K context if loaded as 4bit)
Well it was a scam when I bought it last year
or nearly 4k
8bit works fine too but sometimes it crashes
Yeah, it's really bad right now... and articles are suggesting it might last well into 2027. Imagine what happens if the situation with TW escalates, as some project that happens around 2027-2028... and the implications that would have b/c of the heightened position of AI in the market too.
Great time to buy apple products
micron stopped providing
never!
How's your experience with Devstral 2?
no ram inflation!
just to run gpt-oss-120B
Smart
get 4 lowest config mac minis
Inflation comes preloaded
Sometimes.
those costs like
Nah, I did the math and it's cheaper now since Apple never raised their prices on RAM
ya mac minis actually better priced for value now over random chinese thing off amazon
until their ram prices get bumped to match rest of market
Hi guys I new
They are one of the vendors who locked in RAM prices b/c they ordered a lot... so I don't think it'll affect their upcoming next-gen products coming out early 2026 much either... but we'll see.
Baubleeee
At least current reports show that they will not be raising prices
bauble
Th I forgot to change my username
Get 3 of those and you are set
No
isn't that 1,2k$
Thats what they all say
It was decent. If you are not going for purely vibe coding then it can do so much. Write fairly large snippets for us! It can be your buddy when internet goes down

🙏
not tried anything agentic*
Cool to know, will try that out later
bauble dropping in 3 more minutes
what if it drops now
Now
After this
Hey do u have a cool group like the one in other server or only nerds u talk all the time
Nuh uh
after this
after this
Tf u mean?
Nah thts it
Spawn now bauble
it won't
It's my words against ur
It will now
Told u
Told ya
Are there any cool or humour knowing guys or only nerds on this server?
yay bauble
We all chill ig
@dry kestrel u gotta admit, my calculation is peak
very close

It was just a hunch of yours!
Cmon
Next bauble at 38
Tell me u guess
Let‘s see
fyi bauble have cd time before it start counting msgs
Just annoying that sum people only watch the chat and dont chat
So...
I tried to tell him, @twin grail 🤣
I think if you estimate a time that's 5 min away
then people think they have time to look away and do other stuff adn don't chat in that meantime
so the guess isn't a constant rate of activity
U did ??

Yep. Your timing yesterday was nearly perfect every time.
i can't predict exactly due to texting rate from everyone is different
But i can calculate the time period it will drop
what about now ?
still in cd
what's cd?
cool down
change directory 
told u
gg
heh I kept thinking it meant change dir too
took me forever to realize cooldown

Better luck next time 🤧
dual monitor has it's advantages


I mean sure u can calculate it or predict on the way u feel
I calculated
i prefer do pure calculate
Make up random rules
"Told you"
I had to say it...
hold on, i just realized chrome browser added split screen
Im feeling another bauble soon
you could use split windows
im thinking about getting a portable monitor thing
up to this time for years now ive done all my work only from single mac laptop screen
cant decide if worth it for me tho
Get an iPad
i have an ipad
even just a second hand one
then use the dual thing
it's old tho, don't know if that'll still work
Depending on your budget, nothing is going to come close to the resolution of the iPad
is dual thing through apple ipad good? latency or whatever wise?
It's really good, and you can always use a wired connection
4K in 27' is superior but it's not portable thing
nice guess it couldn't hurt to try, and deffo cheaper if it works since I already own it
Therefore Comet has them too 💅
Just make sure it's compatible https://support.apple.com/en/guide/ipad/ipad1ddf3831/ipados
On your iPad, use Continuity features to share and stream content, answer calls, and more across devices near each other.
pickle
Yeah lol
More like any bauble here
I hope I can get to 50 variants today 
And after this spectacle, I will be retiring from bauble callouts
i went to get some water...
smh
the button shows up before the image most of the time
Imagine living
so I click without knowing which one it is
Just when I was up for coffee
75
so the more you collect, the more difficult it is to track
oh wow...Jen is at 74....but only 47 unique i see
yeah. lots of repeats. and I missed a rare one last night
getting 75 will be nearly impossible
unless the bauble drops at much faster rate
probability of getting 75/75 unique must be very low with only what....72 hours left?
and assuming random bauble type drops - even if you collected every single one from now until the end
there's still 4 days isn't there?
right at 100 hours left. 4 days + 4 hours
HEY GUYS
Of course, I miss a clock buable when I was reading about precision in watch making.
what time and day is it finishing?
I have a new mac air m4 13 inch, but now I regret buying it and am thinking of getting a windows
the 25th at 5pm something ET
Should I swithc
5:40pm on the 25th
still...getting the uniques will be tough
yea 100 hours left makes it a bit more possible i guess
assuming average of 24 baubales a day
100 baubles left
if the chat is active you could get 50+ baubles ever 24 hours
hmm if I can get to 50 unique today
then I'll try for the 75
if not, then I think the number I've collected will be good enough for top 10 and I can just do holiday stuff for the next 4 days
I guess that's the compromise with myself I'll make
baubles
collecting baubles
wdym
it's a coupon so it'd extend by 3 months wouldn't it?
Yep
right
is that confirmed?
not sure why you would need confirmation for something like that
I think they've made it pretty clear
will receive a free
subscription for 3 months
there've been a lot of promotions
He's just bored
which don't add up like that iirc
hmmm
That's external promotions
This is a server event
Berd realistic chance to be top 10
You can check in #🎄┃event-commands
another repeat for you
yep
once u get to about 30-40 unique, i guess odds go down a lot to get uniques
I think they have different drop rates like gachas, and which is chosen is random within that
Top 10 means total numbers of baubles collected or total numbers of variants collected?
total baubles
I don't think the variants actually matter
yea variants don't seem to matter i think, but confirmation would be nice 😛
only if you want to catch em all
I see!
It doesn't matter
but topping the board of having only unique variants is unique 😛
The top ten users with the most baubles collected
No where does it mention variants
doesn't matter that it shouldn't either lol
well it's just a silly statistic
that the total collected also are all unique
cos u can't undo repeats iiuc
I still didn’t get my refund
satisfies the completionist itch. If they didn't tell you how many were in the pool people would ask. Over and over.
Where did you purchase the subscription from?
there should be two prizes maybe
I was very disappointed in the easter egg thing that no one was able to complete their collection
which is why I set out with that as my goal for this
one for top unique, and one for top total
although it's likely the top total also wins unique
being first doesn't matter either - but they tell you who was.
true top 10 same prize
yea top 10
as long as you aren't last +1
lol
1-7 safe zone by a slight margin
Claude Vending
https://vxinstagram.com/reel/DSh6hY7AF41
the top 5 are quite far ahead
True
I got a bauble where I was only one to collect it
not sure if that's ever happened to anyone
Hi
How do you catch bubbles on a mobile? 24/7
just click the button?
Hit the button
Huh
just tap it?
Bubbles??
In the app
baba
It'll Drop soon enough just press collect
Platform?
Wdym
Android/iOS?
???
If Discord is not active, minimized, not in focus, etc
How can I request a refund on iOS?
Go to reportaproblem.apple.com and sign in using the Apple ID associated with your App Store purchase. Navigate to your digital purchase history, and tap "Request a refund.”
you must have discord open to collect a bauble
I email them last month
Any hyperos dev here?
Or LAS LAST
Did you read my message?
last
no
wow
I can open Discord only once-twice in a day
yay new one
:0
400+ APM
How can I request a refund on iOS?
Go to reportaproblem.apple.com and sign in using the Apple ID associated with your App Store purchase. Navigate to your digital purchase history, and tap "Request a refund.”
Weird limitation, use the web browser?
The collect button keeps loading
I cant constantly watch for thread
That's on you...
It's a discrimination 😄
Check this out
https://techforce.gov/
How?
@dry kestrel research quota is probably nerfed to 1 per month.
Costs money to run this :/
lol
I know. Just saying.
Especially after the huge increase in popularity after k2
Only those who have free time and big screens can catch bubble
49" screen helps yea
I’m personally in a movie theater and have the discord open on the big screen 24/7
Bunch of users here have been just using their phones while doing something else
Anyone from Israel I really thank you for your support
although would need 100 unique baubles
Yup i have it open on my phone
Stay on-topic and helpful in each channel. Use #💗│sharing to post tips, content, or cool discoveries.
How to catch bubble message if the Discord app is minimized or closed?
you need discord open
just keep chatting here
and never anywhere else for the next 3-4 days
lol
also don't sleep 😛
Even on Christmas…
Now
especially 😛

hmm
Wsp ssj
How to use other apps on my phone? How to live?

No
after this message
I'm backing you
dang it
Now
baubles are the only life
open it on your computer, when you want to use your phone for something else
now
yes

no to top the leaderboard
Bauble power activate
Spamming won't do anything xD
you're on 20 in like 12 hours lol
we will see about that
We dont spam? We talk
now
U sure?
Time to make 30 alts to keep chatting with kimi k2
Hi
you got Max?
Gautam ik that name
What are the chances “now” is blacklisted
no
Yes
never
Worldguesser?
What
|| please don't ban me moderators ||
Okay he not that gautam
ban
No I am not
the time is bauble
Gautam is a dev of world guessr
I wonder how many times bauble was said from start of event to end
much less
Babubu
You will only get baubles if you have "real" conversations.
labubu?
Babble
any predictions on when ai bubble pops?
Plz pplx team send me all the bubbles to dm so I can in a calm condition click to catch them all when I have free time
is an LLM evaluating this conversation
Gatum is cast
He lying
these are way more real than usual imo lol
furry
Fr
you got Pro now?
Its now or never
Using Gemini on browser cuz app might be nerfed.
we are all prepared
Max might have better tokens etc
They gonna jump that bauble
you may release the goods
or better system prompts
Yeah




