☣︎ larpochka ☣︎
Region: DE
Saturday 13 June 2026 18:46:35 GMT
33519
1794
160
48
Music
Download
Comments
x :
You are not running anything on that level locally 😂
2026-06-14 15:15:32
180
J :
Local inference is ahh unless you have a $10k setup
2026-06-16 15:50:22
1
Prod.danio :
oh yea u got 1.75Tb of vram right ? cuz thats whats needed to run fable 5 locally 😂
2026-06-15 17:47:26
17
⃟⃟⃟⃟⃟⃟🇦 :
holy larpity larp
2026-06-14 20:19:21
42
' :
local models are unusable for any real tasks, anything under 28b js useless
2026-06-13 23:24:51
50
ODD :
“Run ai locally” My cpu:
2026-06-14 22:33:38
18
👽👽👽 gnarp 👽👽👽 :
to run something like fable 5 locally you would need as much as you make a year for each month
2026-06-15 09:36:40
13
Spyrki :
Yes of course we all have a supercomputer 😁😁
2026-06-15 03:14:46
5
ral :
just dont use ai
2026-06-14 13:56:48
2
Coditan Official :
Local models + Coditan = 🔥
2026-06-16 22:03:15
0
_zq2 :
yeah bro i have 600gb of vram for opus 4.7✌️🥀
2026-06-15 06:35:26
6
Vmogus :
government have their own ai which is 2x better than fable
2026-06-16 02:46:32
0
яблоко :
What kind of background is that
2026-06-14 11:21:48
6
Ben :
This doesn’t even make sense on about local models or die, you thinks a 7-13b parameter model is coming close to the performance and output of Fable 5? Bs. For long programming work, local models don’t compete to anything these frontier companies are releasing, for normal day to day work, id rather trust these companies knowing the output will at least be near what i want it to be compared to some shitty fine tuned Llama model
2026-06-14 21:41:29
5
the underdog hedgehog :
realistically, how many people do we need to make an affordable gpu cluster to run a huge model locally?
2026-06-15 20:52:50
0
Karim D :
“Run locally” all good ai models have insane parameters so u can’t run locally…
2026-06-16 19:51:43
1
NurielTal :
Wasted money
2026-06-16 20:36:15
0
salvv :
local models reaching nowhere near opus 4.8 capabilities, and even tho if you would run local models they would be expensive since a good gpu costs 2.8k to run the models.
2026-06-14 13:47:49
0
Mohammadontop :
Im saving money for the nvidias spark 🫂🦥✌️
2026-06-13 18:52:26
3
⃟⃟⃟ :
id like to see you run fable locally lmfao
2026-06-15 17:47:22
5
ironpillar☪️🇹🇷 :
local is crazy expensive for u to have even something close to frontier agentic ai level u would probably need a small server for something like fable 5? def a server at least
2026-06-14 23:15:30
1
erasemymemory :
all a marketing stunt
2026-06-15 03:49:00
0
save :
India just got Alpie ai
2026-06-16 11:47:13
1
Bebbesi :
yeah but local models are not powerfull enough even with a rtx 5090
2026-06-14 15:01:59
1
To see more videos from user @larpu_acc, please go to the Tikwm
homepage.
