this post was submitted on 03 Jun 2026
830 points (99.6% liked)
People Twitter
10147 readers
349 users here now
People tweeting stuff. We allow tweets from anyone.
RULES:
- Mark NSFW content.
- No doxxing people.
- Must be a pic of the tweet or similar. No direct links to the tweet.
- No bullying or international politcs
- Be excellent to each other.
- Provide an archived link to the tweet (or similar) being shown if it's a major figure or a politician. Archive.is the best way.
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I did look it up before I commented at all, and what I was looking at wasn't a good picture, they are pretty close.
my bad. I still am not going to taking any frontier labs word for it, I hope you get that. And for real I was not/am not trying to be a dick, the benchmarks I saw said opus 4.5 was winning out on reasoning, I saw some others that were a lot more mixed.
are you running it? what quant/hardware? how fast is it practically?
I run 27b at q8 with unquantized KV cache and 256k context on two Instinct MI60 GPUs. Definitely the best model that I have been able to run locally at a reasonable speed. 35b generates tokens as fast as you'd expect from any cloud provider. 27b is slower than 35b, of course, but token generation is still faster than my reading speed and suitable with coding agents.
How have I not heard of this line of GPUs?? wth. lot of wattage. I've got a 7900xtx on my desktop and a modded 2080ti with 22gb vram and a 3060ti in my server (my old desktop hardware). I tried 27b dense at q4 or q5 a few weeks ago on my 2080ti and it was painfully slow and I was getting pretty mixed quality results.
32gb of vram is hard to come by.