#💬│general
1 messages · Page 197 of 1
gtx 1650?
Lowkey unusable.
damn
Yes.
Non reasoning models work fine but reasoning models will take ages to respond.
definitely i guess it would have very less token/second speed
this video is fun take on chatgpt
follow form more programming content specially in javascript , python , c++ , java and become skillful coder and developer
Don't forget to LIKE, COMMENT, and SUBSCRIBE for more tech comedy and insightful developer content!
#programming #coding #DeveloperLife#WebDevelopment #UIUXDesign #JavaScript #Python #Jav...
O3 pro be like
Somewhere between 3 and 10 per sec.
ohh i thought buisness pro meant teams one😅
Yes
there's another link for that
You can try ur luck
Lmao
But i don't really rec using this AI
this one is the normal pro too buddy not the teams one( i have normal pro hehe doesnt like it that much due to its context window and nerfed models)
scam and worse than per
yeah i dont generally
true
Tried your luck?
So pplx has unlimited full o3?
i used to when we didnt had o3 on pplx, so used to use o3 from you.com for some research related stuf one month back, nowadays havent opened it much
Wouldn't be so confident about it.
Who's gonna take their time to burn 200+ prompts on o3 in pplx lol
maybe it might be high or medium variant depending on use case like it was for for some model earlier🫠
For science
he meant abt full o3 part, ig rather than unlimited
Ye kinda but per is better anyways
No doubt in that
Got your confirmation or waiting?
Waiting
U can try ur luck
Wouldn’t hurt
I guess they won’t be that broke to sell our emails to russian hackers
Lol
Anyone know of any 3rd party client apps that you can use instead of the official ones?
a few days ago i had to register for an exam, i had a doubt, but their Boucher was big to read, so i gave it to perplexity to read and used the deep research mode even, it gave me wrong information again and again, ended up just not registering for the exam, later found out that perplexity was wrong on this, the information was indeed confusing but instead of admitting that it just gave me a concrete wrong answer
I wan to know how many time of lab pro and research mode I could use per month after I subscribed?
labs = 50 per month
research = unlimited
because all you do is generate images 
Imagine using research
@distant aurora ever tried to fully test out the visual abilities of o4 mini high and o3?
It's shit on you.com and excellent on chatgpt since it can crop, rotate and zoom in manually.
I think it's more than just the lack of tools
That electricity bill test was an excellent example.
Just the base visual detail recognition is insanely good now
Might be. It worked fine on t3 chat but was shit on you.com.
O4 mini high.
They can probably control the amount of visual tokens it can use
I think that's a thing
Ah....
Maybe, idk. I don't know much.
I'm not sure if it's really that. I've seen in docs how visual input is counted as tokens
Taks not visible to me anymore?
I've generated so many images, i think i deserve an artist role
in chatgpt?
the money burner role xd
Yeah
was it available to you before? did any of your scheduled tasks got deleted with it?
I have like 20k images generated on midjournery 😭😭😭 I can't wait for the api to come out
api of MJ ?
No api atm only through discord and web atm
There's ways to self host a api for midjournery but its strictly agienst there tos. They been hinting at getting a api set up for the last couple weeks though. Also they are about to release there first video model
daym
i am looking forward for their video model
MJ has unlimited collection of best images
on earth
if video model executed properly it can come closer to veo 3
This is from the rating party.
Yeah that's mainly because they do rating partys and everything is based of what the user likes the most in terms of style etc for example chatgpt image gen 1 is good but if you use chatgpt for it directly in one chat they do a static seed per chat so its basically impossible to do the same prompt over and in one chat it'll just always give the same output etc
for early stage its gud
Opus... Where are you???
its working fine for me, but not for my alt for some reason.... wbu?
default
What are you doing here? https://cunnyx.com/TheGlobal_Index/status/1932859595019497589
China #1 leader in scientific research
USA #2
China has 8 universities in the top 10
USA has harvard
its better than yesterday
o nice nice
WTHHH
LMAO
yeah unfortunately models are horrible at ascii art
Hm shouldn't be I found o3 to be better at making world maps (only slightly)
probs just other factors playing in
its nice that they at least kept o1
Fr. You.com variety >
Ill be really wary of their subscription plans and avoid anything annual for now though
they silently changed some of the stuff on their pricing page
and is refusing to give me any clarification
Tell me about it.
use wayback machine
its kinda hard to screenshot without making everything really tiny
I use desktop mode on phone.
essentialy they nerfed the limits
Did they put any number there?
no
and thats what I really dislike
gonna be testing that soon after my t3chat credits get used up though
Ikr. I hate this too.
admins refused to answer anything about it as well
Ill be spamming the shit out of o3 though
and hopefully getting an answer
@dry kestrel #💬│general message
what was the prompt?
yeah?
That. Juice score. In case you wanted to know.
mhm
theyve set the thinking limit very low
Bro is in full character.
No information anyways. Deepsearch couldn't find anything on it.
Might wanna spam opus instead.
That's the most expensive model rn.
true
Seems full for now. Don't know about unlimited. Their official claim is 300+ per day.
300 o3 per day

Yeah it was. The help docs say that’s where it’s supposed to be. My take are fine though
Manus AI now supports scheduled tasks too!
@distant aurora https://x.com/ai_for_success/status/1933475474522714365
Any free trial/limits?
No
I doubt anything can even remotely challenge google.
Straight up credit card page to pay 10$
You mean chatllm
same ig
They're
They are the same
Chatllm is product of abacus
Deepagent is a part of chatllm and chatllm is made by abacus
@gloomy cipher did you figure why you got charged even when you had free tokens
Yes...????
Oh
Has anyone here tried giz ai?
Huh
0.25 cents for an image????
Imagen cost you the money right
Because that's probably a different model brother
Img 1 or dall e
Not part of the list
Latest open ai image model idk what it's
It looked beter than the img 1 in perplexity
Generated almost 50k output yesterday no issue
Also i lowered the budget so it doesn't spend
Liquid glue
heklo
Hi - Does anyone also noticed a difference in quality of output model gives when using it directly vs using is via perplexity? I have been experimenting recently by using Claude 4 and Gemini 2.5, and results of using it via Perplexity is quite underwhelming. The results are much better if i use it in Claude chat or Gemini App. This is mostly with respect to coding. One search tells me that Perplexity limits the context window, but even the output under 32K is not comparable.
Are pro models only to be used for search and analysis? Not sure if this was already discussed here and a known thing.
For my usage purposes Perplexity is better than Gemini or free Claude version directly
Pro search is getting pretty close to deep research from a few months ago imo
Just saw coding, mine is finance
They've all improved for me, I use research in a space tho
Might work if you are using paradroid's setup.
Idk what that is, I just use Perplexity 😭
You don't know what good is if you think that is good.
You should try Gemini deep research.
They give 10 free per month.
Anyways.
i hit 600 sites with gemini before 😭
Gemini probably has the best deep research rn, followed closely by Chatgpt.
Anyways, the current addition of O3 was a major improvement to perplexity.
And it's actually good so far.
Gemini is broad and has large source count but OpenAI is thorough smaller source count
It's interesting to see how it goes to previously looked up sources to check again
I found Gemini to be pretty thorough.
But maybe I haven't used it enough cuz I'm free user.
wait gemini.google or google labs gemini?
Gemini.google
I use it when I don't need to go very in-depth lol
Saves the openai requests
Wait until they reduce source again from 10 to 5 lol
ChatGPT for me is when I need to brainstorm something and don't wanna contaminate the thread history lol
I used to have the deepresearch but then it disappeared for some reason
xd so true though
You have to select flash instead of pro.
Then it shows up.
Weird.
You used up all 10?
interesting
You got pro?
Yes
ill try on phone sec
We luck out on the latest AI features being in US, might be another country
pro account
Vpn
Oof name leaked
burner account dw
Doxx fr
Lol not you
oof i see it rip
o me, dw this is my public account
i have my name everywhere
yeah my gmail is supercombogamer and boby LOL
He was talking about himself.
Lol nah
Jay
Yeah that's my online name
o yeah i saw that
Gaymini
i thought hero was but anyways
did anyone try the diffusion model yet?
Better than gpts shit
fr
do u guys think its because im under 18 that it wont let me deepresearch?
Haha hero is from 2017 when i first made my discord acc
Nah
Shouldn't matter
that do not make sense
Wus this
guys i dont think u understand how strict it is
Link
the new diffusion model from gemini on waitlist
i couldnt use apple pay till i was 16
Oh shit
thats how bad it is
i cant use notebook llm either
i asked them about it and they said cause im under 18
BRUH
o well turned out nice
Instagram 🔥
no access for now
Fuck yeah
yeah my brother and i both signed up weeks ago he got in a couple days before me slowly rolling out to people i guess
No comet
I also requested for the free you.com thingy with work email
I hope i get it
Would be fun to have it
Where do you get work email? Same source?
whats freeyou .com?
I work in a company so i have a work domain
You.com is an aggregator
Contact google one support. I had the same for 3-4 days
alright
I wonder what the work emails of star prawns are like.
buy a domain and spin up a email server
Still don't get why they are doing it
imma use all the domains i have
I think there's a verification process
I see.
So it's not automated for sure
Do lmk if you have any luck xd
i gave 2 seperate domains, and imma give my school aswell, so 3x XP is turned on xd
Wish chatllm had a trial
i just did 2 work emails
damn.......
Legit or you had domain ones lying lol
if there is 28k others competiting for the thing, rip me
Haram
Holy cow, 300 per SECOND already, whoa!!!
To avoid that in future
Who has?
guys do u think i ask google about
for notebookllm?
another support agent told me it was probably cause of my age
but like if they really knew the issue they wouldnt have said 'probably'
nice
Perplexity
who hasnt is the real question he should be asking
Flashbacks to yesterday…
Hi CatMatPat
Ah 3 days ago, that incident you mean, or?
@short narwhal do you listen to laufey?
Images aswell?
Guessing they're not exactly your style tho
As far as i saw just py and other coding files
Idk about like listening full on, but I have a few songs on my playlists etc
Good enough
I love her music
Any Red Hood fans here?
Quality Settings in bio!
Follow me for more: @zyon.if
#dccomics #batman #redhood #arkhamknight #batmanarkhamknight #game #edit #4k #fy #foryou
I love Henry moodie.
@timber niche https://www.instagramez.com/reel/DFkloK4NSbY
You are the media now 🫵
Why is perplexity premium not able to generate an EXCEL if it is supposed to use GPT in its processes?
try cvs
Ill do 🙂
It can. Who said it can't? Ask it to make a table. Below the table there's a button to download it as Excel file.
Also, chat gpt has its own set of dedicated tools.
Third parties may or may not have similar tools of their own.
I am doing a calculation for a mortgage,
The system is "vague"
It does not show me all the rows, even if I specify it in the prompt.
True
Bro, the bad boy thirst trap is such a victim card phrase. 😭
What model are you using?
Yeah they deployed a lot of bots on various platforms to reel in more normies
This doesn't mean that the service is actually good and people like it, that's why the number of queries is increasing however.
The service remains dogshit.
Perplexity pro's automatic model
In reasoning mode generates the table
But there are too many mathematical errors
You only have to apply the French amortisation system, it is not a complex calculation.
Bro, lmao.
Use Deepseek R1 or O3.
If you use automatic then you are a disappointment.
Also, make sure whatever you are uploading and making fits under the 32k context limit.
Did Pooplexity replace their gimmick R1-1776 with 0528 already
So, the automatic mode does not work?
It is not able to detect what task I want to do and adapt to it?
Is it just going to use the fastest mode?
did anyone play with o3 pro yet?
kinda
Why is perplexity very slow today
its pretty good
Well people, get ready, World War III continues to escalate until nuclear weapons explode.
Nope.
@languid star @pseudo ferry @gloomy finch https://www.instagramez.com/reel/DK1Q3b-RORP
Follow @pakkarelatable
.
.
#latestnews #latestmemes #planecrash #planecrashindia #planecrash2025 #ahemdabad #ahemdabadnews #newstoday #newsupdate
@distant aurora The prompt I used is a modification of the one posted here lately. R1 answered me that way, it's not a prompt problem, since in Perplexity the R1 works.
You mean the new R1 doesn't follow prompt properly?
Was it a one time thing or did the same issue repeat multiple times?
Do not worry. Besides, the o3 is of no use to me, no matter how much they describe it as "advanced reasoning".
There you saw it in the image
Once
i just received an invite to Comet but when I tried to download the app it kept failing. Now when I try it says "you have downloaded too many times". Is this something I need to wait to reset, or is there some tech support I need to contact? Sorry if this isn't the right place to post - just excited to try out Comet.
more queries, deep research, labs
ask pplx that question
pplx can do that 2 
That’s wild
Gemini really needs memories and custom instructions
I'm another person who scored perplexity pro for free for a year due to having a Samsung phone. I very regularly use both paid ChatGPT and Gemini right now.
I feel like I have a pretty good understanding of those two LLMs and what they're good and bad at. I use ChatGPT for deep research, note taking from transcripts and ideating. I use Gemini for image and video generation, live chat back and forth for brainstorming and if I'm not happy with ChatGPT's results for deep research.
What should I be leaning into for Perplexity? I may very well be at a point of saturation here....I only added Gemini because I needed more storage and it essentially came with it for free vs filling a gap. (Though it is better at some things than ChatGPT)
yeah that would really elevate the expereince
Heya you can use perplexity as a search engine, i would recommend using claude sonnet thinking as your go to model.
Everyday questions using pro search is pretty great
Using labs you can create full dashboards, that are self hosted and shareable.
o since you pay for gemini hmmmmm
Perplexity is sort of an aggregator that provides you a taster of the models from open ai, anthropic, grok etc
yeah true true actually
Also spaces are pretty cool, be sure to check it out
yeah they also have a bunch of templates so you can mess around with that
They essentially provided a container space where you can have custom instructions and upload documents and yout results will be based on that
The live voice feature altho not as good as gpt, it's still usable
Appreciate the insights. I am excited to dig in and play a bit. Also excited for the sharing channel on this discord to see some good use cases. I use AI a TON now -- becoming just the way I think through things.
Only main limitation of perplexity is their 32k context token limit per thread prompt
Yeah the community is awesome here, be sure to check #1124060874785685554
That can help you get some cool prompts from the community
Nice....thank you. Will dig in there for sure.
You can share feedback in #1111765634267742408 if you face issues
You're welcome!
Welcome to the community:)
You're living up to your username...
🤣
why dont i have access to that
what role u need to access it?
Pro
he gave the pro channel this is the general one #👉│feedback-general
bruh im pro but why i no have pro channel 😭
@wraith crow Can you give the role :p
Nah you got it now
thx @wraith crow
they have a backend system for it
how did they get my discord username connected to my perplexity account though?
ngl to you i have no idea lol

thx
i just know they have a system lol
@wraith crow can i also get access to the pro channel? i have pro sub i cant find the discord link
nothing wbu
just finished coding
thx
nice nice, what project u working on currently?
A website for self-learning Ielts
I might change the UI
It's still in beta after all
allow for the side bar to be open and closed
also move the top right text the ielts more to the side
should be right above equally space with the side bar
if side bar is closed can either turn into a logo or keep same
I planning on using Ensemble/Hybrid AES Model for grading mechanic
that one just preference tbh
dw
but should move the text to be evenely above the sidebar
o thats cool

anytime good night\
Nightt
Definitely not for you yet :p
4 hours left
I’ve tried to hit 2am but it always slips
😭
Lmfao
How?
No. I hate myself.
A lot.
It keeps me up.
And the despair keeps me restless cuz I don't have guaranteed solutions to problems.
So I engage myself with distractions.
Another OpenAI employee just dropped a cryptic teaser, clearly referring to GPT-5 and you can tell they’re excited.
The Price of Consciousness
You wondered if the nightmare was ever real- or if you were simply too awake to escape it.
Nietzsche warned:
to dream while seeing clearly
is to live inside a waking abyss.
#Kafka #Nietzsche #ExistentialNightmare
#PhilosophyQuotes #AwakeAndLost #think_and_exist
It better be good
@twin grail Feed that index.html to canvas. You'll surely get impressed!
Hi Everyone
Check out this and mention the company name from which u moved to competitor
I will drop a demo ai call
https://www.linkedin.com/posts/yash-goyal-074a8625a_buildwithvapi-vapi-customerretention-activity-7339386442613145600-EeGQ?utm_source=share&utm_medium=member_android&rcm=ACoAAD_gCNgB7Y2dYmIi76vh47jWkuQXM_TerWY
Want to increase your profits 💸 by up to 90 percentage?
Did you know your profits can grow by 25% to 95% with just a 5% increase in customer retention?
Most companies are focused on acquiring new customers, but quietly losing the old ones — the ones who already trusted you, purchased from you, and are way easier to win back. The problem? No...
@storm kindle legendary lobby
And it was their map.
All of them voted for it.
😭
100 per day o3!
nice
domination tho
i hate it
What do you like except snd?
nah snd is full of sweats with snipers now
and objective modes are too long
i just stick to frontline and tdm
Tdm used to be my favourite.
It got too boring after a certain point.
@gloomy finch https://www.instagramez.com/reel/DJt3BUhyMds
Eternity in Passing Hands
Dostoevsky mourned the soul you gave- to someone who couldn't carry even your name.
Cioran knew the deeper grief: we gift forever
to those just passing through.
#Dostoevsky #Cioran #ExistentialHeartbreak #PhilosophyQuotes #SoulAndSilence #think_and_exist #TragicLove
My feed these days >>
hmm
It's very smart and picks up stuff by itself 
theres so many models baking and cooking now once gtp 5 comes out google will release something or claude on the same tier etc and vise versa its just a moment of who goes first tbh
In my preferences and I'm asking perplexity to list the model used to generate the response and is never matched up with the model i selected. Is perplexity straight out lying to us regarding which models are being used from this selection box?
Getting 40 when I select '03 model it seems like
Sadly models aren’t aware what they specifically are and what they are called
Choosing o3 does use o3, and Claude 4 uses Claude 4 no matter what the model might say
yes to a point the model only knows what it is based of the system prompt being given perplexity prob has a system prompt for all models like you are perplexity etc
Exactly
will i get introuble if i post a github link showing all LLM system prompts even perplexitys?
Heads up guys, there are a lot of servers getting hacked at once so i urge you to check each one you are a part of
Thanks for the clarification everyone! A very odd failure mode, haha.
i sent it in dms
look at your dm
That's actually pretty funny lol
Is the climate impact of AI actually as bad as some people make it out to be
no
o3 pro nano
These aren't leaks lol you can find them in the api web page
Hi
this is 100% not in official documentation
You need to change ur feed algo before it makes u more depressed and hate urself even more bro🫠
Any chance do we have videos in perplexity?
like a video gen? probs not
Like gap image 1 being in perplexity so I guess veo 2
video gen is extremely expensive compared to gpt1 which is already rate limited
likely not happening with a $20 plan
esp if they keep doing these free promos
Does anyone know a way to get stealthwriter subscription for free or for cheap using discount codes?
Hey there, I'm a pro user of perplexity, could I get the comet access or not
no
That code leak will bring more users😭😭
This looks majestic
Iran cooked Israel.
For now atleast.
Israel made a big mistake initiating I think.
Gave the gulfs an excuse to gang on them again.
The study rehabilitates the thinking ability of large models; Apples study is wrong.
W
Israel is always in aggressive mode so consequences are inevitable
Apple can go to hell
What leaderboard
human benchmark
What kinda tasks is it tho
So Apple not only failed to launch a competitive AI, but also failed to downplay others' successes lol
how do i make pplx read an entire github file from the github link itself? i mean all the files in it.
I thought that was not possible but after some back and forth and reasoning , it literally read everything i wasn't able to make it to do so eariler
@halcyon coral
Pain… 💔
Original knight falling animation by @blendreams
#meme #memes #sad #wholesome #ww2
What timezone is icelavaman in?
🤔
Wasn't this study only viable for slow old models
I heard o3 pro crushed this argument
Woah thanks pplx
They really need to bring memories and tasks to the app
You can see these tasks in your library once it does them though
Ah okay
At least it can recall memories on mobile
Insane censoring by me
Fbi level redaction
@storm kindle should I spend or save these?
is this ios 26 beta (with glass thingy turned off?)
It's prolly not out yet
It is
But I don't think mat is using it.
There's plenty of videos on twitter regarding how it didn't turn out as expected.
the beta is out. i think the beta 1 drops the same day as wwdc
Ah okay
maybe iran was on the verge of fully completing nuclear
I’m not on 26 yet
Ask again next month
They be funding Hezbollah too.
lol ok
can anyone plz answer this?
Do the files fit within the 32k context limit?
Depends on Russia and Saudi Arabia.
i am not "uploading" any files. it should be able to read all the files from github itself
heh russia is a bit occupied
Yeah. Rag still consumes tokens.
i mean it did access all of them once after it had finished giving me multiple revisons of code. i thought it would be better to start a new chat and let it read everything first, but now it just calls the search tool and says i can't read more stuff or something
but its capable of doing so because i saw it do it multiple times in the previous chats
idk the trigger for that yet. thats the issue
Wdym "people are free"
So the illegitimate country that carries out a full fledged genocide and attacking another nation on a whim, deserves no pushback and has to be free
I am confusion lol
I have pro subscription. How do I get a pro channel ?
most civilians there are actually heavily against the regime there
Yeah but israel doing what it's doing isn't going to change the regime, just going to get more people (civilians) killed + Israel don't give a fuck about anyone else but their asses
yes but they hate israel even more and while they do they wouldnt want be invaded like what you did in iraq and killed millions in the name of peace
The word philistine comes from judeo words from Judaism
Keep feeding yourself the bs
I don't want to argue with someone who is so in on themselves
Hahahaha ok whatever you say mr smartypants
Fact doesn't care about feelings
which model?
And feelings don't care about the facts
Muh feelings T-T
Well eg. O4 mini high for plus
You need deep research about everything before knowing and saying something.
context window by default is 200K token but iam not sure if chatgpt has a limit on their website
What is ragging
so as long you dont exceel 200k context thats fine btw history also counts in the context windows
dayum thats a big nerf
Prob, not specified
Not gpt
Nah that's ChatGPT
If someone contributes to society I will deeply respect you. Otherwise Idc about your feelings.
https://openai.com/chatgpt/pricing/ go look for yourself
how do you define that lol i was working with this mathmatician recently who is palastinian and he introduced the SD Math program to the world
Maybe 10-20
oh so you decide who is useless and useful
agreed. but idk why it "feels" "different" from evey other ai model i have used
Google cooking something after their recent model is taken deprecated
Oh I remembered palestinian leader stealing billions of dollars and going to dictatorship Qatar.
are you hearing yourself bro? do you know how redicalize you sound? but ig its not radical because its not targeting the western society
Hahaha how to change the topic
Yes.
Jews contribute most to nobel prizes
By most i means majority
So I am in their side
Victim mentality doesn't work here.
Lololol talk about yourself
Keep crying
stalin whom a lot of athiest sees him as their god killed millions and starved tens of million , according to your logic and reasoning we should hate all athiest now?
But idk i have plus and it can definitely read a smaller book well
such as?
how big is the book in words count
Cheers mat
@short narwhal improved research is out as per announcements.
Yeah I saw
Research is coding!!
#📢│announcements message @green crystal
DAMN
They really have beefed it
Opus maybe
👀
Hopefully the daily DR limits remain the same
wasn't able to do this before
I like itt
Seems to be
i personally didn't notice any major improvements. has it rolled out to everyone or still rolling out in phases?
Dope asf
Same. I don't see any significant difference.
Though there are less mistakes.
Definitely has more output token limit seeing this
App even breaks lol
I suggest you to not judge it by source count alone. Focus on the quality of responses too.
yeah same as before for me
clearly you haven't seen the ChatGPT app
Lol haven't used it a lot yes
It's fine for me. The response is shortened on the app.
spend if you cant collect all shards within time limit
theres still like 20 days left
I don't think ak 47 will be removed after 20 days cuz 56 amax is still there.
https://chatgpt.com/share/681e495b-9828-800c-9a95-da259a57a803 see if you can open this up in the app
Is there an open in app option on the website?
noo clue
Hi
Hey 👋
Is there any badge etc here for the perplexity pro users?
you have the pro role
You already have the pro role
That's the "badge"
Did it break for you mat
You should see this in the desktop apps
Mac app straight up doesn't load it for a good bit and on windows my PC starts screaming
amax doesnt show a time limit on it
Is there time limit for ak? I don't see any.
iirc i saw it saying 20 days or smth to grab it
https://www.perplexity.ai/search/conduct-a-comprehensive-search-cZxnfQfSQhW1xhQmrSBswQ?0=d#0 this seems to be a bit more in-depth
wait lemme see
@cobalt vale @dry kestrel @green crystal the responses aren't as much lazy but there's still mistakes in battery capacity/processor which I'll get to later.
Will use some other query to test source count limits.
Lmfao
Meanwhile my phone browser loading it just fine.
hm i doesnt say it on that page
i was probably seeing things being deprived of sleep for 30 hours
I feel like the game's getting kinda boring cuz of the shitty matchmaking.
@storm kindle https://youtu.be/e0Z6d4IA2PI?t=19m17s
Play RP GRAND NOW: https://rpgrand.com/?utm_source=youtube&utm_medium=integration&utm_campaign=OXY
.
Use promocode “OXY” to get free RARE ITEM!!
.
Subscribe
Other platforms:
• Instagram -- https://www.instagram.com/oxyyyyy/
discord -- https://discord.gg/CXMQyh7C4u
tiktok -- tiktok.com/@oxyontiktokk
•Discord personal -- #oxy5229
•I...
imma save it to watch later
Okk
Ok ok I should have gotten it before maybe because i thought i left the server and rejoin
are u a straight female or a gay male
wtf
ist not gonna be anything else
what u mean ist?
Bots are getting weirder these days
@distant aurora you gonna test new research?
Why doesn't perplexity pro work
If it's the one we've been promised? Lol
you are attracting them
from PPLX?
well
iam using it rn but it doesnt show the source number
did they remove it ? 😂
nvm
59 sources 👀
fresh new bought grapes? XDDDD and 5g very expensive emmental cheese? XD
cat kibble and dried mealworms resoaked with water few days old from earlier meals?
u mean labs is the new researc htool
iam testing the output limit it claims to have written 10'000 word in one go but after counting the words it seem to be 7700 🤔 so i guess around 7000 word is the limit of output? which by itself is actually a lot way way more than normal search can handle usually you have to convince normal search to release 500-1000 words per msg
https://www.perplexity.ai/search/can-you-make-a-paper-about-cau-8UkuwHBdQSu8flzqiubCFw can others work with this thread?
or just me?
like is it a shared project?
@gloomy cipher
I uploaded a recording. I'll test more.
Perplexity guide please
good
I ain't talking about you
In general human population are some of the dumbest races in this planet. They are like leech who destroy everything on this planet.
They are so selfish and only tiny minority of they contribute to society and planet.
No wonder so many things are going endangered.
Everyone lacks critical thinking and logic.
Tried. Doesn't work. Too many humans. We need population purge.
Animals endangered, Climate changes happens because of us.
Global warming is happening because of us.
Humans are selfish and shitty.
I don't give a shit about humans unless they contribute to society.
The majority of them are shitty and selfish. Only a tiny fraction contribute to society.
how can you be logical and critical thinking if your solution to a wound is cutting the limbs
and i would like to know of your contributes to society please iam interested
By tiny percent I meant it less than 0.1
Me ?
yes that was metaphor but yes thats what i understand from your view
you aknowledge societies issue but instead of solving them with logic and critical thinking you prefer annihilation
I feed stray dogs and animals every night instead of giving it to beggars. On my birthday I donated to blind cottage where blind peeps are shunned by society since they are unable to see and do jobs. I go there every year to donate money and food.
I prefer annihilation because human are leeches
I got access to Comet yesterday. I must say that so far it feels like a breath of fresh air. Nice clean user experience, and thoughtful integration of Perplexity AI features.
They destroy everything they touch kinda like the story of Midas
I do have some achievement but I don't like to share here.
Works for me. Make sure CORS is enabled, everything else should be exactly like this with the model name being the exception
I need to try this later. I'm a brand new Obsidian user, and I was wondering if something like this existed.
I prefer dogs and cats over human beings
i thank you for your contribution but
- why would giving food to either beggar or dogs be a bad thing? in my humble opinion i see both as good deeds and contribution.
🙏 Comet with Notion might be fun too
but thats a personal preference what if someone else preferce helping a pregnant woman over a begga or a sick person or an old individual , these silly questions are not important thats not what we should be discussing what we should be happy about is the fact that someone is contributing wether its for a dog/cat/beggar/oldman/pregnant women.
- instead of worrying ourselves about which group in needs should get food lets worry about not commiting evil doings
how you guys use obsidian? i tried and returned to vscode+markdown plugin + dendron plugin
i found obsidian lacks rich amount of 3rd party extensions
you can connect obsidian with api which i did but idk how it works i couldnt find a proper guide
and you can get free api in open router if your job is easy, if its heavy reasoning tho you can benefit from the open ai free token 5$ trick
I've used obsidian in a very basic way, gets the job done for me
Been thinking about switching to something cloud based just for the accessibility of it 🤷
obsidian is cloud based
connect the vault folder to your google drive
free cloud service
yeah I have it connected to my iCloud
But sometimes I wish I used something like Notion just to use it in the browser as well
Where?
oh grasps mouth as i knowledge iam facing an apple user
Erh how's the thought process 
What does it do
note
notes
Oh I'm stupid lol
Though there are some errors like this
Can't we just do it normally on the app lol
jay
Why does it need cloud
yeah but obsidian is more powerful
ik you dont take any notes
markdown and plugins for example
to share it between moba and pc
Ah okay that makes more sense
And you can add ai api huh
That's cool
With plugins, yep
What's the search and thought process like?
Nice
Yeah I write in binary in my brain
yea to ask question ex if you have 10 page note you can ask what your notes says about specific things
im not sure where in the spectrum im in
i basically use vscode w/ plugins for note taking
So like notebook llm
non binary? 🤨
Reasoning is still so basic like how it was
Smh
I've been using cursor to use some md files lately ngl...
yes
You are erratic today haha
It did some math in python I'm guessing.
what do you mean? i do have mood swings if thats what you meant
Yeah but there's no reasoning between searches
That's what the deep research preview did
Ah... that.
Fair.
Did you get the free 3 month pro
i also tried but i missed the extensions vscode can provides...
Fair, seems like you have your own workflows :p
I have my ways
It's running on sonar I think.
Yeah
That would explain the inaccuracies in specs.
Or atleast let o3 do the searching and sonar the final reasoning and summary
I think that would be less tokens than o3 doing everything
I'm still figuring it out.
Atm it still feels like pplx research does wide searches all over the place and attempts to reason over everything in the end
@gloomy cipher where r u from
how 
This is what it gave me
when i said it need to merge 2 of this project with the main one being the newer one, just add the missing function
from this
With this
It took o3 pro 17 mintues
And it gave me that
Noice
this question out of nowhere 👀
northern iraq
what if you ask it to use framework like react/svelte?
What are your interior motives 🤨
what are you trying to do i thought you were studying korean not english
or do you need to have an ielts certifiate aswel? for visa
Yes i am
i'm studying korean
Nope
i got my ielts cert
scored total 8.0 overall
amazing, academic?
Yup
Good luck
thanks ❤️
iam quiet freaked out actually i dont think i can get a high score 😔
u got this
dw
any advice? how did you practice beforehand
Reading or listening ?
ngl seeing someone with gen z ahh slang and 'u' gets 8 brand score does raise up my hope

yea or in general
umm for reading
did you pay for any tutor or any course?
it depends on ur hobbies
i myself found out i kinda like reading novels or any science researches
is where u can find numerous of research from various aspect
it also gave u the necessary knowledges for the reading part
harry potter ( eng ver )
or any kind on novels
iam reading novel, news outlet a lot
Nope
For listening
I got a 7.5 years back after just watching some YouTube videos on understanding the format
i usually listen to the listening test at x1,75 speed
The reason why is when u got familiar with a higher speed
the ability to hear is more advanced
but 1.25 should be good
The speaking one was the hardest


