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

DeepSeek V4 Pro 0813

openrouter.ai

Post reply on HN
Print view View original

451–460 of 493 posts

Re: DeepSeek V4 Pro 0813

#451

Post by LeBit »

gnunez wrote: post #262
bigmadshoe wrote: post #198

Earlier quoted context omitted.

Why are people giving these n=1 comparisons like they mean anything? The worst offender is that pelican guy. These are non-deterministic systems and a single trial should not update your priors much at all. Of course it's significant that your response had a bug and took four times longer, but if you're only going to try once, this isn't real science, it's just vibes.

I don’t understand why people are calling these transformers models non-deterministic? Are you referring to the temperature parameter? I haven’t played with transformer internals in a while but my understanding is that if the temperature is fixed at a value where the top logit is always picked, then because they weights are fixed, the exact same input should produce the exact same output. Am I missing something?

You could answer your own question really, really quickly.

Re: DeepSeek V4 Pro 0813

#453

Post by neuroticnews25 »

jiscariot wrote: post #432
simonw wrote: post #136

Nice bicycle chain, the little basket with a fish didn't show up in the right place: https://tools.simonwillison.net/markdown-svg-renderer#url=ht...

Wonder if we'll ever see optimizations for pelican riding a bicycle svg, make it in to model training runs.

Wondering about this in simonw pelican threads is part of the tradition.

Re: DeepSeek V4 Pro 0813

#454

Post by kaeluka »

spijdar wrote: post #130
svachalek wrote: post #91

Earlier quoted context omitted.

Maybe it's me but I don't see how DS Flash is better than GLM at all, much less by a huge gap. I'd probably protest less against Fable and Opus being put at the same level than many would, but there's no denying the two models are a very different experience from each other. I guess where I'm going is no one should pick a model by the benchmarks.

I'm not the most LLM-savvy person around, and I'm not gonna say I've put a ton of effort into practically compared these open models. But, a month or two ago I did do some "practical evaluates" testing GLM 5.2 versus DSv4 (flash/pro) with OpenCode's subscription with some late 80s Unix clone-type work, and this jives with my experience. GLM ended up being far slower, and far more expensive, for approximately the same…

Interesting!

GLM 5.2 is slower for sure (although they offer a fast version), and it's more expensive. But in my experience, it's universally better than Deepseek V4-flash-0731. Don't get me wrong, the new Flash version is amazing.

But the use cases I have looked at are about source code understanding, bug finding, etc. - GLM 5.2 is clearly better.

I think by using some prompt engineering, you will probably be able to close this gap, but some extra work is needed.

And I'll say it again: the new Flash version is amazing. I love it. That level of intelligence for the price is unprecedented, and the fact that it's open weights and runs locally makes me genuinely happy.

Re: DeepSeek V4 Pro 0813

#455

Post by dominotw »

disgruntledphd2 wrote: post #409
dominotw wrote: post #221

Earlier quoted context omitted.

i think thats BS that harness has equal weight. most of intellegice is still coming from training data not from RL. so how is 'coevolved harness' equal weight.

> most of intellegice is still coming from training data not from RL. For coding specifically, I'm not sure this is still true. Given the heavy use of RL to improve coding performance, I'd expect the harness to be important as it defines what tools the model is rewarded for using.

i belive they looked at the traces and they were still legible ( RL traces should be gibberish)

Re: DeepSeek V4 Pro 0813

#457

Post by cyanydeez »

f6v wrote: post #433
cyanydeez wrote: post #144

Earlier quoted context omitted.

the breadth and width of the universe of oneshot challenges are all arbitrary. It's unsurprising different workflows oneshot better than others. All the more reason to favor local models under your control, as once you find that sweet spot model, no one can change it, upgrade it, align it, take it down or otherwise harm the time investment you made it making it your own. I can't really believe no one understands, aft…

But I don’t want to manage a local model…

Likr sll infrastructure, once its setup and your flow is set, the only thingin the way is FOMO.

Re: DeepSeek V4 Pro 0813

#459

Post by m00dy »

DeepSeek is moving to peak/off-peak API pricing. Off-peak rates are 50% of peak rates.

Peak: 01:00–04:00 UTC and 06:00–10:00 UTC Off-peak: all other hours

New pricing takes effect August 16, 2026 at 16:00 UTC.

Model Period Cache hit Cache miss Output (input / 1M) (input / 1M) (/ 1M)

deepseek-v4-flash Off-peak $0.007 $0.22 $0.66

deepseek-v4-flash Peak $0.014 $0.44 $1.32

deepseek-v4-pro Off-peak $0.022 $0.66 $1.98

deepseek-v4-pro Peak $0.044 $1.32 $3.96

For batchable workloads, scheduling outside those two UTC windows cuts token costs in half.

Re: DeepSeek V4 Pro 0813

#460

Post by rng-concern »

zero0529 wrote: post #397
jeswin wrote: post #371

Earlier quoted context omitted.

For a while now, I've found pelican rendering to be an unreliable metric for LLM ability - and most people know it. Yet, somehow it gets upvoted to the very top of every new model discussion.

Well it is just a bit of fun I think. However, I also think an AGI or an extremely capable model approaching AGI would be able to paint a pelican on a bicycle fairly easily. So in that way it is a good metric.

I agree it's fun, no argument there.

However, it's no longer a good metric, as "drawing svg pelicans" is now showing up too much in the training data, so is not proof of generalization.

Post reply on HN