Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Qwen3.8-Flash-Next

qwen.ai

Post reply on HN
Print view View original

101–110 of 246 posts

Re: Qwen3.8-Flash-Next

#101

Post by mattalex »

jrflo wrote: post #39
rohansood15 wrote: post #31

Earlier quoted context omitted.

This is a good counter argument. But you have to note that this is after OpenAI cut Luna costs by 80%. If you compare launch pricing, Qwen probably comes out ahead on a cost-performance basis.

The luna cost cuts were real though, not a one time promotion or something, due to some optimization (probably distillation?) that openai did.

You assume that openai's inference is profitable and that they aren't just trying to bolster revenue before their IPO.

The only indication that openai is profitable comes from openai (whom I wouldn't trust with any statement, especially when it comes to profitability).

In fact there is evidence that inference is not profitable simply because the rate of losses doesn't seem to reduce as revenue increases: if inference had great margins, we would expect that as revenues increase, the amount of spend on training reduces as a fraction of total expenses. Since the loss-making fixed costs shrink as a fraction compared to the profitable inference, we should expect profitability to rise with total revenue.

However, all leaks of openai's numbers seem to suggest the opposite: as revenues increase so do the losses.

Re: Qwen3.8-Flash-Next

#102

Post by redox99 »

Squarex wrote: post #12
rohansood15 wrote: post #9

Didn't expect it to beat 3.8 27B so cleanly. Opus 4.6 Max self-hosted at 30 tok/s on a 5k Macbook in Aug 2026. The LLM timelines are crazy.

I don't like these comparisons. Sure it is impressive, but it does not have a world knowledge of larger models. It has most of theirs intelligence.

At 125B + 51B I'd expect it to have some degree of world knowledge, clearly in the middle between small models like qwen 27B, and huge trillion parameter models.

Re: Qwen3.8-Flash-Next

#103

Post by hluska »

peterleiser wrote: post #97
redrix wrote: post #53

Earlier quoted context omitted.

You’re absolutely right to be hopeful. Three honest possibilities, and I’ll be straight with you about each: 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 irrationall…

> Three honest possibilities, and I’ll be straight with you about each This. I don't know if the "honest answer" phrasing is part of the system prompt or alignment, but when people say "honestly" all the time I start wondering how honest they're being.

Thankfully most people have better reading skills than that.

Re: Qwen3.8-Flash-Next

#104

Post by stymaar »

Imustaskforhelp wrote: post #41

Pelican: https://gist.github.com/SerJaimeLannister/8fdef9c00175da0ca6... 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 certainl…

> I think that Longcat was the first one who used it

Wasn't it introduced by Gemma?

Re: Qwen3.8-Flash-Next

#105

Post by redox99 »

LaurensBER wrote: post #22
Squarex wrote: post #12

Earlier quoted context omitted.

I don't like these comparisons. Sure it is impressive, but it does not have a world knowledge of larger models. It has most of theirs intelligence.

If/when we can get larger context this will mostly be mitigated by these smaller models being able to search the internet. Self-learning/improving would be even better but that's still a long way to go.

Search results suck because the web sucks these days. The big models from OpenAI/Anthropic have every book in existence baked into them

Re: Qwen3.8-Flash-Next

#106

Post by hluska »

mattalex wrote: post #101
jrflo wrote: post #39

Earlier quoted context omitted.

The luna cost cuts were real though, not a one time promotion or something, due to some optimization (probably distillation?) that openai did.

You assume that openai's inference is profitable and that they aren't just trying to bolster revenue before their IPO. The only indication that openai is profitable comes from openai (whom I wouldn't trust with any statement, especially when it comes to profitability). In fact there is evidence that inference is not profitable simply because the rate of losses doesn't seem to reduce as revenue increases: if inference…

I don’t pay OpenAI’s bills - I pay what they charge me. Their cost accounting isn’t relevant to a user.

Re: Qwen3.8-Flash-Next

#107

Post by schopra909 »

Can someone explain the intuition behind the en-gram idea? I know DeepSeek published a paper about it a few months ago and the Gemma models have a lightweight version of it; but it hasn’t clicked for me yet

Re: Qwen3.8-Flash-Next

#108

Post by serf »

dist-epoch wrote: post #47
rohansood15 wrote: post #27

Earlier quoted context omitted.

For world knowledge, you'd want it to find and reference the source material to be sure. At that point, it doesn't matter if the knowledge is embedded.

World knowledge also means knowing the various algorithms and ways particular programming problems are solved. You can't search what you don't even know exists.

>You can't search what you don't even know exists.

that's not really entirely true -- one can google for "fast pathfinding' and stumble upon A-star , all that had to be queried was the intent/desire.

a lot of smaller agentic models and a lot of harnesses live on that premise.

Re: Qwen3.8-Flash-Next

#109

Post by overgard »

xatnys wrote: post #83
lnenad wrote: post #21

Adding to my homelab stack, hopefully it doesn't overthink like the little model. Actually, hoping it thinks a bit less. Wait actually I'm really praying it reasons a bit more directly. But wait, I'm really sure that it must be a bit better.

Did you observe the model overthinking on practical tasks? While 3.8 does think a lot on xhigh I've found that it really depends on the task. On one-shot prompts that are usually the first to be posted during new releases it will tend to spend a lot more time thinking than doing. In other words the more open ended a problem space becomes, the more Qwen will tend to second-guess itself. Conversely I've found that it c…

Yeah, I ran into an overthinking loop with it a couple days ago on a task that shouldn't have been that hard. (It's kind of interesting to watch the internal conversation happening with it). Overall I'm impressed with it, but setting the /effort to medium is what you usually want (it defaults to xhigh). I do wonder if I had made it write out a plan if I would have avoided that though.

Re: Qwen3.8-Flash-Next

#110

Post by Casteil »

lnenad wrote: post #21

Adding to my homelab stack, hopefully it doesn't overthink like the little model. Actually, hoping it thinks a bit less. Wait actually I'm really praying it reasons a bit more directly. But wait, I'm really sure that it must be a bit better.

Yep.. for 'general purpose' use I found qwen3.8:27b to be disappointing due to overthinking. It's brutal especially considering how slow it is compared to MoE variants. It often overthinks to the magnitude of ~10x the tokens vs a ~4x faster gemma4:26b-a3b.

As a result, qwen3.8 will churn over a prompt often for 5-10 minutes while gemma4 regularly finishes the same prompt in under 20 seconds, while giving a consistent and accurate response in my favorite test case. Qwen3.8, despite churning like that, often misses with an inaccurate answer.

Obviously, 'YMMV' depending on your use case... just sharing my two cents.

Post reply on HN