upvote

points

by zozbot23416 hours ago |
comments
upvoteby jonaustin8 hours ago|
next
[-]
And shout-out to Qwen if they release 122b -- Jeff Barr's original Gemma 4 tweet said they'd release a ~122b, then it got redacted :(
reply
upvoteby bertili16 hours ago|
prev|
next
[-]
Is there any source for these claims?
reply
upvoteby zozbot23416 hours ago|
parent|
next
[-]
https://x.com/ChujieZheng/status/2039909917323383036 is the pre-release poll they did. ~397B was not a listed choice and plenty of people took it as a signal that it might not be up for release.
reply
upvoteby anonova16 hours ago|
parent|
prev|
[-]
A Qwen research member had a poll on X asking what Qwen 3.6 sizes people wanted to see:

https://x.com/ChujieZheng/status/2039909917323383036

Likely to drive engagement, but the poll excluded the large model size.

reply
upvoteby stingraycharles16 hours ago|
prev|
next
[-]
397A17B = 397B total weights, 17B per expert?
reply
upvoteby zackangelo16 hours ago|
parent|
next
[-]
17b per token. So when you’re generating a single stream of text (“decoding”) 17b parameters are active.

If you’re decoding multiple streams, it will be 17b per stream (some tokens will use the same expert, so there is some overlap).

When the model is ingesting the prompt (“prefilling”) it’s looking at many tokens at once, so the number of active parameters will be larger.

reply
upvoteby wongarsu16 hours ago|
parent|
prev|
next
[-]
397B params, 17B activated at the same time

Those 17B might be split among multiple experts that are activated simultaneously

reply
upvoteby littlestymaar16 hours ago|
parent|
prev|
[-]
That's not how it works. Many people get confused by the “expert” naming, when in reality the key part of the original name “sparse mixture of experts” is sparse.

Experts are just chunks of each layers MLP that are only partially activated by each token, there are thousands of “experts” in such a model (for Qwen3-30BA3, it was 48 layers x 128 “experts” per layer with only 8 active at each token)

reply
upvoteby kylehotchkiss13 hours ago|
prev|
[-]
How many people/hackernews can run a 397b param model at home? Probably like 20-30.
reply
upvoteby jubilanti13 hours ago|
parent|
next
[-]
You can rent a cloud H200 with 140GB VRAM in a server with 256GB system ram for $3-4/hr.
reply
upvoteby jmb992 hours ago|
parent|
prev|
next
[-]
I’ve mentioned this as an option in other discussions, but if you don’t care that much about tok/sec, 4x Xeon E7-8890 v4s with 1TB of DDR3 in a supermicro X10QBi will run a 397b model for <$2k (probably closer to $1500). Power use is pretty high per token but the entry price cannot be beat.

Full (non-quantized, non-distilled) DeepSeek runs at 1-2 tok/sec. A model half the size would probably be a little faster. This is also only with the basic NUMA functionality that was in llama.cpp a few months ago, I know they’ve added more interesting distribution mechanisms recently that I haven’t had a chance to test yet.

reply
upvoteby Nav_Panel7 hours ago|
parent|
prev|
next
[-]
The point is that open weights turns puts inference on the open market, so if your model is actually good and providers want to serve it, it will drive costs down and inference speeds up. Like Cerebras running Qwen 3 235B Instruct at 1.4k tps for cheaper than Claude Haiku (let that tps number sink in for a second. For reference, Claude Opus runs ~30-40 tps, Claude Haiku at ~60. Several orders of magnitude difference). As a company developing models, it means you can't easily capture the inference margins even though I believe you get a small kickback from the providers.

So I understand why they wouldn't want to go open weight, but on the other hand, open weight wins you popularity/sentiment if the model is any good, researchers (both academic and other labs) working on your stuff, etc etc. Local-first usage is only part of the story here. My guess is Qwen 3.5 was successful enough that now they want to start reaping the profits. Unfortunately most of Qwen 3.5's success is because it's heavily (and successfully!) optimized for extremely long-context usage on heavily constrained VRAM (i.e. local) systems, as a result of its DeltaNet attention layers.

reply
upvoteby blurbleblurble2 hours ago|
parent|
prev|
next
[-]
It only has 17b active params, it's a mixture of experts model. So probably a lot more people than you realize!
reply
upvoteby kridsdale313 hours ago|
parent|
prev|
next
[-]
I can (barely, but sustainably) run Q3.5 397B on my Mac Studio with 256GB unified. It cost $10,000 but that's well within reach for most people who are here, I expect.
reply
upvoteby qlm13 hours ago|
parent|
next
[-]
Hacker News moment
reply
upvoteby toxik13 hours ago|
parent|
prev|
next
[-]
$10k is well outside my budget for frivolous computer purchases.
reply
upvoteby zozbot2349 hours ago|
parent|
next
[-]
It would be plenty in-budget if the software part of local AI was a bit more full-featured than it is at present. I want stuff like SSD offload for cold expert weights and/or for saved/cached KV-context, dynamic context sizing, NPU use for prefill, distributed inference over the network, etc. etc. to all be things that just work for most users, without them having to set anything up in an overly error-prone way. The system should not just explode when someone tries to run something slightly larger; it should undergo graceful degradation and let them figure out where the reasonable limits are.
reply
upvoteby bdangubic12 hours ago|
parent|
prev|
next
[-]
99.97% of HN users are nodding… :)
reply
upvoteby hparadiz11 hours ago|
parent|
[-]
There are way too many good uses of these models for local that I fully expect a standard workstation 10 years from now to start at 128GB of RAM and have at least a workstation inference device.
reply
upvoteby bdangubic10 hours ago|
parent|
[-]
or if you believe a lot of HN crowd we are in AI bubble and in 10 years inference will be dirt cheap when all of this crashes and we have all this hardware in data centers and it won't make any sense to run monster workstations at home (I work 128GB M4 but not run inference, just too many electron apps running at the same time...) :)
reply
upvoteby bigiain7 hours ago|
parent|
next
[-]
> I work 128GB M4 but not run inference, just too many electron apps running at the same time.

This is somewhat depressing - needing a couple of thousand bucks worth of ram just to run your chat app and code/text editor and API doco tool and forum app and notetaking app all at the same time...

reply
upvoteby hparadiz10 hours ago|
parent|
prev|
[-]
Inference will be dirt cheap for things like coding but you'll want much more compute for architectural planning, personal assistants with persistent real time "thinking / memory", as well as real time multimedia. I could put 10 M4s to work right now and it won't be enough for what I've been cooking.
reply
upvoteby stefs9 hours ago|
parent|
prev|
[-]
yeah, but if you really really wanted to and/or your livelyhood depended on it, you probably could afford it.
reply
upvoteby lpnam02011 hours ago|
parent|
prev|
next
[-]
In where I am living, 10k USD is a little more than 3 years worth of rent, for a relatively new and convenient 2 bedroom apartment.
reply
upvoteby gbgarbeb34 minutes ago|
parent|
[-]
$277 a month for a two bedroom is literally 6-10% of what someone in the SF Bagholder Area pays.

Either you're in Africa, southeast Asia or south/central Amarica.

How do you even afford internet?

reply
upvoteby SlavikCA13 hours ago|
parent|
prev|
next
[-]
I'm running it on my Intel Xeon W5 with 256GB of DDR5 and Nvidia 72GB VRAM. Paid $7-8k for this system. Probably cost twice as much now.

Using UD-IQ4_NL quants.

Getting 13 t/s. Using it with thinking disabled.

reply
upvoteby GrayShade1 hours ago|
parent|
[-]
I get 20 t/s on the UD-Q6_K_XL quant, Radeon 6800 XT.
reply
upvoteby rwmj12 hours ago|
parent|
prev|
next
[-]
For some reason you were being downvoted but I enjoy hearing how people are running open weights models at home (NOT in the cloud), and what kind of hardware they need, even if it's out of my price range.
reply
upvoteby kylehotchkiss8 hours ago|
parent|
prev|
[-]
you have proved my point
reply
upvoteby r-w13 hours ago|
parent|
prev|
next
[-]
OpenRouter.
reply
upvoteby mistercheese13 hours ago|
parent|
next
[-]
Yeah I think there’s benefits to third-party providers being able to run the large models and have stronger guarantees about ZDR and knowing where they are hosted! So Open Weights for even the large models we can’t personally serve on our laptops is still useful.
reply
upvoteby parsimo201013 hours ago|
parent|
prev|
[-]
If you're running it from OpenRouter, you might as well use Qwen3.6 Plus. You don't need to be picky about a particular model size of 3.6. If you just want the 397b version to save money, just pick a cheaper model like M2.7.
reply
upvoteby bitbckt11 hours ago|
parent|
prev|
next
[-]
I'm running it on dual DGX Sparks.
reply
upvoteby adrian_b9 hours ago|
parent|
prev|
next
[-]
The 397B model can be run at home with the weights stored on an SSD (or on 2 SSDs, for double throughput).

Probably too slow for chat, but usable as a coding assistant.

reply
upvoteby xienze8 hours ago|
parent|
[-]
I think you have that backwards. Agentic coding is way more demanding than simple chat. The request/response loops (tool calling) are much tighter and more numerous, and the context is waaaaay bigger in general.
reply
upvoteby ydj11 hours ago|
parent|
prev|
next
[-]
Running the mxfp4 unsloth quant of qwen3.5-397b-a17b, I get 40 tps prefill, 20tps decode.

AMD threadripper pro 9965WX, 256gb ddr5 5600, rtx 4090.

reply
upvoteby stavros13 hours ago|
parent|
prev|
[-]
It doesn't matter how many can run it now, it's about freedom. Having a large open weights model available allows you to do things you can't do with closed models.
reply