Running local models on an M4 with 24GB memory

#1

Post by shintoist » Sun, May 10, 2026, 11:09 PM UTC

Running local models on an M4 with 24GB memory
jola.dev

Re: Running local models on an M4 with 24GB memory

#2

Post by sbassi » Sun, May 10, 2026, 11:33 PM UTC

A useful data to know about this setup is how many tokens/sec generates.

Re: Running local models on an M4 with 24GB memory

#3

Post by JBorrow » Sun, May 10, 2026, 11:40 PM UTC

sbassi wrote: post #2

A useful data to know about this setup is how many tokens/sec generates.

It’s started in TFA

Re: Running local models on an M4 with 24GB memory

#4

Post by NDlurker » Sun, May 10, 2026, 11:50 PM UTC

JBorrow wrote: post #3

sbassi wrote: post #2

A useful data to know about this setup is how many tokens/sec generates.

It’s started in TFA

You can't expect someone to read 4 paragraphs into an article before commenting

Re: Running local models on an M4 with 24GB memory

#5

Post by NBJack » Sun, May 10, 2026, 11:55 PM UTC

I'm puzzled. The M4, as far as I know, doesn't have 24GB. Did the author mean a M40?

Re: Running local models on an M4 with 24GB memory

#6

Post by spoonyvoid7 » Sun, May 10, 2026, 11:57 PM UTC

NBJack wrote: post #5

I'm puzzled. The M4, as far as I know, doesn't have 24GB. Did the author mean a M40?

M4 = M4 Macbook Pro

Re: Running local models on an M4 with 24GB memory

#7

Post by sertsa » Sun, May 10, 2026, 11:58 PM UTC

NBJack wrote: post #5

I'm puzzled. The M4, as far as I know, doesn't have 24GB. Did the author mean a M40?

M4 Mac Mini w/24GB sitting right here on my desk.

Re: Running local models on an M4 with 24GB memory

#8

Post by tra3 » Sun, May 10, 2026, 11:58 PM UTC

NBJack wrote: post #5

I'm puzzled. The M4, as far as I know, doesn't have 24GB. Did the author mean a M40?

There’s definitely an option with 24 gigs of ram: https://support.apple.com/en-ca/121552

Re: Running local models on an M4 with 24GB memory

#9

Post by kennywinker » Mon, May 11, 2026, 12:00 AM UTC

NDlurker wrote: post #4

JBorrow wrote: post #3

Earlier quoted context omitted.

It’s started in TFA

You can't expect someone to read 4 paragraphs into an article before commenting

@grok is this true?

Re: Running local models on an M4 with 24GB memory

#10

Post by canpan » Mon, May 11, 2026, 12:01 AM UTC

Recent models (Qwen 3.6 and Gemma) can really do coding locally. Feels like SOTA from maybe a year ago? But you would want about 32-40GB total memory. 24GB is just a bit short of that. A gaming PC with 16GB graphics card and 32GB RAM brings you very close to a usable coding system.