The prompt processing times I've heard about have put me off wanting to go that high with memory on the M series (hoping that changes for the M5 series though). What's the average and longest times you've had to wait when using opencode? Has any improvements to mlx helped in that regard?
replyThe M5 ultra series is supposed to have some big gains around prompt processing - something like 3-4x from what I've read. I'm tempted to swap out my m4 mini that I'm using for this kind of stuff right now!
replyWow, Kimi K2.5 runs on a single M3 Ultra with 512 GB RAM?
replyCan you share more info about quants or whatever is relevant? That's super interesting, since it's such a capable model.
How's the inference speed? What was the price? I'm guessing you can fit the entire model without quantization?
replyExcellent. Thanks for the info!
reply