@gigaqian: Alibaba just released a SUPERPOWERED AI model that’s small enough to run locally! Qwen 3.8 27b released a few hours ago, and groups like Unsloth are already optimizing quantized versions. This model is extraordinary with tool use, computer interaction, coding, and agentic workflows. This video will help explain why it’s so strong, what makes it different, and how you can start using it today! #breakingnews #ai #localai #china #gigaqian

GigaQian
GigaQian
Open In TikTok:
Region: US
Friday 14 August 2026 23:03:14 GMT
20571
1546
77
192

Music

Download

Comments

oficinamk3
Marcos Machado :
Hi GigaQian! Thanks a lot for the tip. I installed Qwen 3.8 27B on my RTX 3090 24GB, and it ran incredibly fast. The local system architecture analyses have been getting the green light when cross-checked against the paid version of GPT! That's awesome! Kudos.
2026-08-16 04:03:34
1
geebuscripes
Tower Lord :
Yes, look at what China just did, but don’t be dismissive about the Epstein files. It’s not funny.
2026-08-15 21:40:38
11
freddybigpants
Freddybigpants🇬🇧 :
Is China really just messing with US here? In a kind of “enjoy your useless data centres!” Kind of way?
2026-08-15 05:25:17
39
daniela.dogioiu0
Daniela Dogioiu :
Qwen 3.8 is great
2026-08-21 20:37:14
0
accaaccd
゚゙゚゙゚ :
Which AI has the best overall reviews right now?
2026-08-15 10:34:51
2
iboynesh
Young money :
is it in hugging face
2026-08-15 05:56:02
3
mccullochyo
McCullochYO :
what about a 4GB GPU? I'm poor
2026-08-15 00:15:49
3
fuji_hikari
Seraph :
I'm waiting for China to start pouring out cheap DDR5 and their new CUDA-compatible GPUs 😁 Nvidia is already terrified, RAM sharks literally faked up 😁
2026-08-15 13:10:22
9
hootie5150
hootie5150 :
I don’t really understand any of this, but wow!!! It tickles my brain when super smart people talk. The comment section is spectacular for me to try and understand 🤓
2026-08-16 17:41:26
7
manure674
manure :
Do you know a model I can run on 8gb vram?
2026-08-16 03:46:19
0
beanie_man5
Darrel123 :
Patiently waiting for nvidia to release the RTX Spark so I can really run great LLM’s locally and not pay for tokens.
2026-08-15 04:38:12
6
alimorfeq
A 77 :
How about i9 32gb ram , 5060 16gb vram?
2026-08-16 13:55:46
0
augiejc
AugieC :
So, I think I finally want to build a local box for this. No more paying the Claude subscription.
2026-08-15 19:34:20
1
cratosza
Cratos Za :
I have a 5070 32gb ram is there anything I can do to run Qwen 3.8
2026-08-15 01:06:26
1
williamquinn96
William Quinn :
brilliant 👌
2026-08-15 01:48:51
1
aratinamaze
Lauren :
I’m confused help I want to understand
2026-08-16 06:30:58
3
mrlocal123454
MR.LOCAL :
Great information, thanks for taking the time 👍😁💙
2026-08-15 00:51:53
1
mrlocal123454
MR.LOCAL :
Are these possible to integrate into Drones?
2026-08-15 00:52:57
2
ash_bash_888
ash_bash_888 :
It’s my birthday 🎁
2026-08-15 01:27:23
2
markooscha
Маркуша :
First 😀
2026-08-14 23:05:13
1
xxxsinxxx
sin :
Is there anything like Moores law for AI optimization?
2026-08-16 18:45:43
0
_dig_dug
Doug :
I need a better computer so I can run this.
2026-08-15 05:43:44
1
chef.foxxy
ChefKENA | Dinner Rescue🦊🩷🌈 :
this model runs without token completely and no monthly subscription??? Question what is the minimum size computer you recommend 8GB 16GB or 32GB of ram?
2026-08-18 13:10:23
2
fergusstevens
fergusstevens :
Thanks Mat! Looking forward to finding a bit of time and hardware to give it a go.
2026-08-15 04:48:52
1
l192846104763
. :
Smarter than 2 trillion kimi k3??
2026-08-15 20:27:14
2
To see more videos from user @gigaqian, please go to the Tikwm homepage.

Other Videos


About