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

Qwen 3.8

twitter.com

Post reply on HN
Print view View original

281–290 of 793 posts

Re: Qwen 3.8

#281

Post by rapind »

anonzzzies wrote: post #197
chewz wrote: post #184

Earlier quoted context omitted.

Have you actually used Opus 4.8 in Claude Code? It takes way too long to do any practical task on higher thinking levels due to over-engineering. And I am not the only one complaining. Lots of people downgrade to Opus 4.6 exactly for this reason. Opus 4.8 training works well for agentic work. Not for code harness. EDIT: ``` stronger on coding and raw capability but can be more argumentative, verbose, and costly. Reli…

Works extremely well for us. I never know what other people are doing when we read these stories.

Yeah I found 4.7 and 4.8 to be downgrades from 4.6. I don't know if it was the model or just Anthropic's scaling issues though TBH. I found working with the Claude Code max (20x) sub would work awesome in between new models. A week before up to 2 weeks after, it would go to crap, dumber, slower, outages, harness churn, etc.

I'm finding the same with ChatGPT recently since the 5.6 release. Not as bad though, but sluggishness at times, harness churn (creating bugs and crashed), and occasional availability issues that cause me to downgrade to 5.5.

It's gotten to the point where I dread a new model release from these companies because it's guaranteed to be disruptive! I assume the pay per use API is less impacted.

Re: Qwen 3.8

#282

Post by versteegen »

kmarc wrote: post #127
onlyrealcuzzo wrote: post #98

Earlier quoted context omitted.

Who do you buy DeepSeek from? I bought it through OpenRouter and used it with Pi agent. The model was good, but there appeared to be a pricing glitch or something, because it burned through $50 in under an hour on pretty trivial stuff. Pi agent claimed it only used like $1. OpenRouter claimed differently and said I used all $50.

I can highly recommend OpenCode Go. I use it from pi.dev as well through the OpenCode Go $10 subscription ($5 first month). Used more than 20M tokens at a cost of ~$20 (up to $60 is included in the $5 plan) Out of which deepseek pro had ~200 messages which is around 1.5M tokens (10+M cached)

BTW the quotas for Go have very recently changed, now only $15 for some models instead of $60. Which is not actually a difference for DS4 Pro, because they lowered the token pricing 4x at the same time (to match the change in official pricing from DeepSeek months ago)

Re: Qwen 3.8

#283

Post by andsoitis »

JKCalhoun wrote: post #256
gardnr wrote: post #6

Earlier quoted context omitted.

It's hard to say what their motivation is. The Chinese firms seem to be working hard to commoditize intelligence which may be the most effective way to debase American frontier labs. And yeah: it also happens to be really good for humanity.

Xi Pitches China as Leader of New Global AI Order, Challenging US Dominance: https://www.reuters.com/world/asia-pacific/chinas-xi-promote...

> and pledged to help developing nations build AI capabilities

Data centers?

Re: Qwen 3.8

#285

Post by andsoitis »

ricardobayes wrote: post #96
gardnr wrote: post #6

Earlier quoted context omitted.

It's hard to say what their motivation is. The Chinese firms seem to be working hard to commoditize intelligence which may be the most effective way to debase American frontier labs. And yeah: it also happens to be really good for humanity.

I'm paraphasing but the Chinese premier said recently AI should be seen as a common good that should benefit everyone.

Then why doesn’t he give it away for free?

Re: Qwen 3.8

#288

Post by Levitz »

WarmWash wrote: post #250
piguin wrote: post #199

Earlier quoted context omitted.

My own experience is that the vast majority of programmers have experience with one model and maybe some short usage of earlier models from a competing choice but want to be using the model with the highest popularity and reputation. I've worked with people who actually had to test multiple choices for their team who didn't understand why they were pressured to select Claude for programmer morale.

When iOS users are given an Android phone, they complain about how awful of an OS it is. In reality, they just aren't used to it.

What exactly is there to get used to? Do people really use each model very differently? There's a learning curve as in everything, but there is nothing that comes to mind for me when I use codex as opposed to claude

Re: Qwen 3.8

#289

Post by apitman »

adrian_b wrote: post #258
smnplk wrote: post #148

Do this giant open-weight models have less active params and could be run on consumer hardware or no ?

Any open-weights model that has ever been published can be run on consumer hardware, even on a mini-PC. The right question is which is the speed that can be achieved on a given hardware and whether it is high enough for the model to be useful. Until now, the speeds reported for running big LLMs with the weights stored on SSDs have ranged from as low as a token every 10 seconds or so, to as high as a few tokens per se…

I can only imagine what that does to the poor SSD

Re: Qwen 3.8

#290

Post by voxgen »

taosx wrote: post #274
gchamonlive wrote: post #245

Earlier quoted context omitted.

It's useless to talk about models and harnesses without context and method. Depending on how you use the model and what the model is used for, experience may vary drastically. Also, different models with different harnesses require different approaches. I've been using https://gitlab.com/gabriel.chamon/orisun which is my own simplified methodology, for coding web apps in python and elixir and have been very successfu…

I'm not sure about "useless" but from my experience agentic coding leads to death by a thousand cuts for all projects I've seen so far. Small decisions missed in a codebase that leads to degradation in correctness, reliability and performance. At some point it only takes one engineer to be careless, others skipping PR because they are AI generated...

It's a solvable problem if you're willing to throw more tokens at it. Frontier models have gotten very good at cleaning up their own messes. You just need the right skills/loops, and to stick to models that consistently follow instructions (i.e. GPT-5.5/GPT-5.6-Sol).
Post reply on HN