> trained at just 1/9 the cost of Qwen3.7-Plus, while outperforming it across the board
:)
It is very relevant and for a certain group of us, far more impactful to our work the next month(s) than any blog post could be.
I wonder if I could get this running through vLLM on 6x Nvidia L4 - the 3.6 worked great on 4 cards but sadly TP6 just isn’t a thing and I don’t have 8 cards available, maybe it’s gonna be okay with like TP2 and MTP. I have no idea at this time, probably need to test out what even might be possible.
This "next" release adds a new concept, first public release with n-grams, I think. And it's in a MoE size that is likely to be very fast and cheap to serve (faster than 27b for sure). It's also well suited for inference on alternative compute (i.e. sparks, macs, etc) so it's relevant to local users.
Qwen's advances do (currently) have merit.
> trained at just 1/9 the cost of Qwen3.7-Plus, while outperforming it across the board
Opus 4.6 Max self-hosted at 30 tok/s on a 5k Macbook in Aug 2026. The LLM timelines are crazy.
Self-learning/improving would be even better but that's still a long way to go.
Also, heating my home during the winter is nice.
Oh, also, I use llamacpp with --reasoning-budget; very simple way to move on.
1. It overthinks — Just like the previous iteration. High confidence. 2. It doesn’t overthink — Improvement from the last model for your use case. Regression for others. 3. It sometimes overthinks — Best case all around. A feature, not an impairment.
One final thing worth mentioning: (I made myself irrationally angry writing this)
Qwen 3.8 flash: $0.16 / $0.47
Compared to
Deepseek 0723: $0.03 / $0.075
(units in USD/m tok)
Original error: llama.cpp does not support this GGUF's model architecture ('qwen4exp')
Edit : Saw the pull request, should arrive soon enough https://github.com/ggml-org/llama.cpp/pull/27742
It's new arch demo for future Qwen 4 family, but (as I understand) training recipe/data is same as any other 3.8 model.
6B active params helps around the memory bandwidth constraints, but a 128GB box can probably run the Q3/Q4 quants fairly easily with a decent context size. This might actually be better for strix users than 27B, which was already very good.
Aside from the pelican, I am sort of impressed by the fact that things are going the way in terms of really impressive small models.
Also I love how this uses N-gram embedding. I think that Longcat was the first one who used it (I submitted that submission on hackernews because I really just loved the idea of it that I understood), I am certainly more interested in local LLM models and its interesting how they are utilizing new architectures to do some really impressive optimizations!
(Do note that I created it using a free rate limited end-point that I found on the huggingface space section: https://victor-chat-with-qwen3-8-flash-next.hf.space)
You can't search what you don't even know exists.
How much memory does this translate to and what quantization (if any) were applied?
Luna is $0.20 / $1.20 vs $0.16 / $0.47 with Qwen.
Given the timing, I think they A. shat their pants since Deepseek flash just came out with insane pricing before the price hikes, and B. Anthropic is really struggling in model tiers below opus.
It was smart for them to cut prices regardless of whether they had 80% efficiency gains or not
whwhyb•1h ago