128
Qwen 3.8 Max (2.4T-a95b) and 27B open weights being released next week (lemmus.org)
submitted 2 weeks ago* (last edited 2 weeks ago) by BeefAndPoultry@lemmus.org to c/localllama@sh.itjust.works
16 comments fedilink hide all child comments

https://qwen.ai/blog?id=qwen3.8

The model weights will be open-sourced on Hugging Face and ModelScope next week — stay tuned.

I'm hoping they also release a new 35b a3b, for us VRAM poors, a new 9b would also be great!

top 16 comments
sorted by: hot top controversial new old
[-] the_wonderfool@piefed.social 8 points 2 weeks ago

Can't wait for the new 27b. The old 3.6 is still such a powerful small model. I also agree for the 35b a3b. I thankfully refreshed my PC for LLMs before prices skyrocketed, so I can run comfortably the 27b one Q8, but sometimes having the pure speed of the a3b is very useful.

[-] melfie@lemmy.zip 3 points 2 weeks ago

It’s not clear from the wording whether the 27B model is dropping next week as well, or whether they’re just announcing that there will be a 27B model.

[-] BeefAndPoultry@lemmus.org 3 points 2 weeks ago

true, it's not perfectly clear

also I just saw this

[-] fluxx@mander.xyz 3 points 1 week ago

Given they have not released 27b or any smaller model with Qwen3.7, I'm not holding my breath for 3.8. They also haven't said anything public about it.

[-] baguettefish@discuss.tchncs.de 10 points 1 week ago

OP's image is literally them saying something public about it

[-] fluxx@mander.xyz 5 points 1 week ago

No, they said it for 3.7 too, but then never released it, that's what I meant

[-] baguettefish@discuss.tchncs.de 3 points 1 week ago

Ah that makes sense.

[-] SirDimples@programming.dev 2 points 2 weeks ago

I hope their decision to opensource Qwen3.8-Max means this version is going to be another open-weights frontier contender 👍

[-] MIXEDUNIVERS@discuss.tchncs.de 2 points 2 weeks ago

i have a 6700xt with 12gb vram. i need a usable model. 9b is barly usable because of token speed. But non the less i'm exited

[-] melfie@lemmy.zip 4 points 1 week ago* (last edited 1 week ago)

The setup described in this Codacus video is what got me started getting over 100k context with decent speeds on my RTX 3070 that I’ve been using daily for the last month: https://m.youtube.com/watch?v=0AqpaFm11oI.

The TheTom fork of llama.cpp adds asymmetric TurboQuant support that allows k at tq4 and v at tq2, which allows squeezing in more KV cache without quality loss. The REAP version of the MoE model also works just fine while further reducing the model size.

The TheTom fork doesn’t have pre-built container images, so it’s necessary to build your own (the ROCm Dockerfile in the devops directory).

If anyone does want to use the fork, I just recently built the ROCm image myself and ran into an issue where the fork’s Dockerfile was pulling a UI build package from Huggingface that doesn’t exist anymore, so I swapped in the latest ROCm Dockerfile from upstream that now builds the UI from source, and that worked fine.

[-] Schilling2304@thelemmy.club 4 points 2 weeks ago* (last edited 2 weeks ago)

With which quantization ? I have a 16GB GPU, running Qwen3.5 9b UD_Q8_K_XL basically max out the VRAM usage. Maybe a MOE model fits you. Gemma 4 e4b is one with a total 8b.

And what is the token speed you are getting ?

[-] brucethemoose@lemmy.world 4 points 1 week ago* (last edited 1 week ago)

How much CPU RAM do you have?

At 12GB you should be using a sparse MoE with offloaded experts. There are a few to choose from, but it just depends on how much spare RAM you have.

[-] SirDimples@programming.dev 3 points 2 weeks ago

I only have 8gb vram and can run 35b-a3b perfectly (30+ tps), have you tried 35b?

[-] TheLegendaryAssholeOfJushinLiger@sh.itjust.works 1 points 1 week ago

Yeah do what these guys are suggesting and do the MOE with the experts offset. It really works well.

[-] BeefAndPoultry@lemmus.org 1 points 2 weeks ago* (last edited 2 weeks ago)

have you tried Qwen 3.6 35b a3b? check my guide, it's still relevant to you just with different numbers because you have 12GB

https://lemmus.org/post/24235317

this post was submitted on 03 Aug 2026
128 points (100.0% liked)

LocalLLaMA

5043 readers
23 users here now

Welcome to LocalLLaMA! Here we discuss running and developing machine learning models at home. Lets explore cutting edge open source neural network technology together.

Get support from the community! Ask questions, share prompts, discuss benchmarks, get hyped at the latest and greatest model releases! Enjoy talking about our awesome hobby.

As ambassadors of the self-hosting machine learning community, we strive to support each other and share our enthusiasm in a positive constructive way.

Rules:

Rule 1 - No harassment or personal character attacks of community members. I.E no namecalling, no generalizing entire groups of people that make up our community, no baseless personal insults.

Rule 2 - No comparing artificial intelligence/machine learning models to cryptocurrency. I.E no comparing the usefulness of models to that of NFTs, no comparing the resource usage required to train a model is anything close to maintaining a blockchain/ mining for crypto, no implying its just a fad/bubble that will leave people with nothing of value when it burst.

Rule 3 - No comparing artificial intelligence/machine learning to simple text prediction algorithms. I.E statements such as "llms are basically just simple text predictions like what your phone keyboard autocorrect uses, and they're still using the same algorithms since <over 10 years ago>.

Rule 4 - No implying that models are devoid of purpose or potential for enriching peoples lives.

founded 3 years ago
MODERATORS
pax@sh.itjust.works
noneabove1182@sh.itjust.works
Smokeydope@lemmy.world
MonsterBug@sh.itjust.works