I'm running Qwen 4 on a Macbook Pro Max M1 with 64gb and it runs really well
2026-11-19 19:19:41
2
Tom :
What are the token per second
2026-11-19 14:58:55
2
Tillar Technologies LLC :
I'm waiting bruh
2026-11-20 09:55:02
0
AR MADMAN :
Lies
2026-11-20 09:03:48
0
Walter in Vermont :
GPT-6 Astra says that it’s choice for a local system to be able to run itself on is the Apple M series, M5Ultra maxed out at 500GB for now, but better yet the M7Ultra maxed out in late 2027. It thinks that by the time the M7Ultra maxed out with memory is available that the software will also be ready. Gemini agrees. We are about to hit the crossover point of both hardware and software for local models. Cost, about $20K to $32K so something you could have on your desktop. My first significant computer in 1979 was $3,500 which in 2026 money is $16,271.19 but that was a 1MHz Z-80 with 4KB of RAM and a tape drive. The MacStudios are a steal today.
2026-11-20 03:55:29
0
Ben H :
average AI users don't code. they have a subscription to chat GPT or the equivalent to make cat videos or ask a quick search question. local models are completely capable of that
2026-11-19 19:45:52
1
The Bearded Taz :
@NetworkChuck is the ACTUAL creator for this video
2026-11-19 19:59:15
1
aarongibbs695 :
Nothing new here this has always been true of any AI model. The question is are you willing to wait for 5 minutes after sending a simple prompt and wait again 3-4 seconds per word while it writes the response
2026-11-20 17:06:39
0
To see more videos from user @user707346657741, please go to the Tikwm
homepage.