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

Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

qwen.ai

Post reply on HN
Print view View original

351–360 of 400 posts

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#351

Post by zrn900 »

jxmesth wrote: post #182
alex7o wrote: post #71

Ok I find it funny that people compare models and are like, opus 4.7 is SOTA and is much better etc, but I have used glm 5.1 (I assume this comes form them training on both opus and codex) for things opus couldn't do and have seen it make better code, haven't tried the qwen max series but I have seen the local 122b model do smarter more correct things based on docs than opus so yes benchmarks are one thing but realit…

The only reason I'm stuck with Claude and Chatgpt is because of their tool calling. They do have some pretty useful features like skills etc. I've tried using qwen and deepseek but they can't even output documents. How are you guys handling documents and excels with these tools? I'd love to switch tbh.

You can just use Cline in VSCode to get most of the tooling you need - it works with all models. Including Xiaomi's new Mimo with 1m context window and blazing fast speed. It's much cheaper than Claude's biggest plan and with much, much more quota.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#352

Post by chillfox »

slopinthebag wrote: post #225
Aurornis wrote: post #43

Earlier quoted context omitted.

For serious work, the difference between spending $10/month and $100/month is not even worth considering for most professional developers. There are exceptions like students and people in very low income countries, but I’m always confused by developers with in careers where six figure salaries are normal who are going cheap on tools. I find even the SOTA models to be far away from trustworthy for anything beyond thro…

$100 / month will get you rate limited to much to rely on with the Claude plans. People still report getting rate limited on the $200 / plan. Also not everyone wants to use Claude Code, so if they're paying API pricing it's more likely thousands of dollars a month. If you can get the same results by spending a fraction of that, why wouldn't you?

Managing context size and efficient token usage is a skill.

I have an Anthropic API key for work, and if I use sonnet/opus all day for agent coding, it ends up costing about ~$25.

I am going to need more cpu/ram to run multiple agents in parallel to spend much more than that.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#353

Post by deaux »

strel0k1 wrote: post #314
enochthered wrote: post #304

Earlier quoted context omitted.

Anthropic limiting Claude subs to Claude code is what pushed me away in the end because I wanted to keep using Pi.

Just sign up for an AWS account and use the Anthropic models through Bedrock which Pi can use.

What advantage are you saying this has compared to just directly going through the Anthropic provider? They are the same price.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#354

Post by ihowlatthemoon »

greyskull wrote: post #324
ihowlatthemoon wrote: post #320

Earlier quoted context omitted.

I run a setup similar to yours and I've had the best results with Qwen3.5 27B. Specifically the Q4_K_M variant. https://unsloth.ai/docs/models/qwen3.5 I use llama-server that comes with llama.cpp instead of using ollama. Here are the exact settings I use. llama-server -ngl 99 -c 192072 -fa on --cache-type-k q4_0 --cache-type-v q4_0 --host 0.0.0.0 --sleep-idle-seconds 300 -m Qwen3.5-27B-Q4_K_M.gguf

Thanks, I'll have to continue experimenting. I just ran this model Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL and it works, but if gemini is to be believed this is saturating too much VRAM to use for chat context. How did you land on that model? Hard to tell if I should be a) going to 3.5, b) going to fewer parameters, c) going to a different quantization/variant. I didn't consider those other flags either, cool. Are you having…

35B-A3B means it's a MoE model with 35B total parameters but with only 3B active at once. The one I use is the 27B dense model. Usually, dense models give better responses, but are slower than the MoE. With your 4090, you should be able to get about 50 tok/s with the dense model, which is more than enough for practical use.

If you want to keep using the same model, these settings worked for me.

llama-server -ngl 99 -c 262144 -fa on --cache-type-k q4_0 --cache-type-v q4_0 --host 0.0.0.0 --sleep-idle-seconds 300 -m Qwen3.5-35B-A3B-UD-Q4_K_XL.gguf

For the harness, I use pi (https://pi.dev/). And sometimes, I use the Roo Code plugin for VS Code. (https://roocode.com/)

I prefer simplicity in my tooling, so I can understand them easier. But you might have better luck with other harnesses.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#355

Post by 3836293648 »

culi wrote: post #99
ninjahawk1 wrote: post #29

The way to develop in this space seems to be to give away free stuff, get your name out there, then make everything proprietary. I hope they still continue releasing open weights. The day no one releases open weights is a sad day for humanity. Normal people won’t own their own compute if that ever happens.

I think that's an overgeneralization. We've seen all the American models be closed and proprietary from the start. Meanwhile the non-American (especially the Chinese ones) have been open since the start. In fact they often go the opposite direction. Many Chinese models started off proprietary and then were later opened up (like many of the larger Qwen models)

GPT started off open? They just closed before anyone else even joined the space

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#356

Post by lossolo »

elorant wrote: post #350
Barrin92 wrote: post #303

Earlier quoted context omitted.

>This is obviously a strategic move at a national level. no it isn't. That's the kind of thing people say who've never worked in the Chinese software ecosystem. It's how the Chinese internet has worked for 20+ years. The Chinese market is so large and competition is so rabid that every company basically throws as much free stuff at consumers as they can to gain users. Entrepreneurs don't think about "grand strategic…

If this was true then they’d build services around those models and provide those for free or vastly cheaper than western competition. But that’s not what they’re doing. Instead they’re giving away the entire model for free. And by the way, Qwen isn’t build from some random entrepreneur who’s trying to solve the cold start problem, but from Alibaba which is a fucking behemoth. And surprisingly of course none of these…

> And by the way, Qwen isn’t build from some random entrepreneur who’s trying to solve the cold start problem, but from Alibaba which is a fucking behemoth.

DeepSeek, Kimi, GLM, etc. are not built by behemoths, and they are free. You do not understand China's culture and market.

> And surprisingly of course none of these models answer uncomfortable questions about China’s past.

Download the GLM 5.1 weights and ask about Tiananmen Square, it will tell you what happened.

You are viewing China through a Western lens. I used to do the same many years ago, but after traveling to China many times, I realized that was a mistake.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#357

Post by StilesCrisis »

ulfw wrote: post #330
operatingthetan wrote: post #262

Earlier quoted context omitted.

It seems like people can't even agree which SOTA model is best at any given moment anymore, so yeah I think it's just subjective at this point.

AI is a complete commodity One model can replace another at any given moment in time. It's NOT a winner-takes-all industry and hence none of the lofty valuations make sense. the AI bubble burst will be epic and make us all poorer. Yay

Staying power is probably the most important factor, which is why I'm thinking Google eventually takes the crown.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#358

Post by vidarh »

alex7o wrote: post #71

Ok I find it funny that people compare models and are like, opus 4.7 is SOTA and is much better etc, but I have used glm 5.1 (I assume this comes form them training on both opus and codex) for things opus couldn't do and have seen it make better code, haven't tried the qwen max series but I have seen the local 122b model do smarter more correct things based on docs than opus so yes benchmarks are one thing but realit…

I don't find GLM 5.1 beating Opus personally, but I do think it is good enough to consider it part of the SOTA pack at this point. It feels like it needs more time and tokens to achieve things, but that's okay - it's so much cheaper per token.

If Qwen3.6-Max is up there as well, it will be very interesting.

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#359

Post by vidarh »

mikenew wrote: post #261
alex7o wrote: post #71

Ok I find it funny that people compare models and are like, opus 4.7 is SOTA and is much better etc, but I have used glm 5.1 (I assume this comes form them training on both opus and codex) for things opus couldn't do and have seen it make better code, haven't tried the qwen max series but I have seen the local 122b model do smarter more correct things based on docs than opus so yes benchmarks are one thing but realit…

GLM 5.1 was the model that made me feel like the Chinese models had truly caught up. I cancelled my Claude Max subscription and genuinely have not missed it at all. Some people seem to agree and some don't, but I think that indicates we're just down to your specific domain and usage patterns rather than the SOTA models being objectively better like they clearly used to be.

I feel like it's Sonnet level for implementation, but not matching up to Opus for planning.

But I agree it's close enough that it's worth using heavily. I've not cancelled my Claude Max subscription, but I've added a z.ai subscription...

Re: Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

#360

Post by salad-tycoon »

tasuki wrote: post #213
culi wrote: post #104

Earlier quoted context omitted.

That's fair but those days seem so long gone now. Also the Chinese models aren't following a typical American SaaS playbook which relies on free/cheap proprietary software for early growth. They are not just publishing their weights but also their code and often even publishing papers in Open Access journals to explicitly highlight what methods and advancements were made to accomplish their results

Any idea why they do that?

If the question is why Chinese models are contributing to open source and sharing of information, I don’t pretend to know the rationale but I think it’s because it’s an economic war.

I think the Chinese models have to be more open to increase trust as everyone is worried they are feeding their very essence/soul into a Chinese copying machine.

Also China wants there to be viable competitors so that US can’t just dominate a potentially very important field. It’s a challenge to a unipolar USA dominated world.

Also it helps to spur Chinese companies in the all important microchip industry which is controlled by a very small number of companies at various steps in the supply chain.

I wonder too if it allows them to hold an ace in their hand as well in terms of threat/power for negotiations. As in, they can cause the whole house of cards to crumble, an economic nuclear weapon so to speak.

Finally, there is a certain amount of prestige involved too. China can compete or even win at a very complicated game. They use it to increase national pride and to project their advancing power status to other nations.

Anyways, just my thoughts. Interested in others thoughts.

Post reply on HN