Live data from Hacker News

DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

artificialanalysis.ai

241–250 of 342 posts

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#241

It’s exciting that a model scoring this high is dirt cheap. It’s also so inefficient, when they release the full performance numbers it’s not going to be good. One example, it takes about 3.6x more tokens to finish the same work as Gemini Flash 3.6.

you are correct. but in my experience -- not benchmarks -- g flash 3.6 is SO BAD for coding. I'm using all vendors all day and gemini is the worst by far. I built my own semi-deterministic orchestrator for coding agents.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#243

Already beat Luna on price/task, by about 2x: https://artificialanalysis.ai/models/deepseek-v4-flash?intel...

If only they would let you opt out of training use, it might actually be a viable option.

The training use is probably responsible for this improvement tho.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#244

Earlier quoted context omitted.

> They announced it already I meant something I could download and run.

Isn't Reasonix [0] DeepSeek's own harness? Or, are they building a new one? [0] https://reasonix.io/

DS were hiring harness Eng team on x a while ago.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#245
post #164

Earlier quoted context omitted.

How often are you asking an LLM about this? You know you can just google it, right? I have to admit it rarely comes up in the coding tasks I usually give to LLMs.

Surely you realize he or she is not literally looking for the answer to that but rather pointing out the censorship.

Of course. But it's a bit like putting an image of the tank man on a circuit board if you order from china (which I have seen discussed, never found someone who actually did it). Most people don't do it, not because they are for or against censorship, but because they are just there to order circuit boards.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#246
post #111

Earlier quoted context omitted.

I'm not looking forward to it, them being strapped for resources provides a huge incentive to develop and release these smaller models. Even if they'd still release their models once they are able to comfortably service all potential customers via their cloud, running them locally would be almost impossible due to their size.

Chinese are pretty pragmatic. Even if they produce more expensive chips and memory, they are still going to focus on value. But who knows, maybe India will step up again like they did for Y2K 3 decades ago.

[flagged]

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#247

Earlier quoted context omitted.

File hosting is pretty cheap. Egress traffic is about $90/TB in the cloud, but around $1/TB in the real world. Storage is in the realm of $5/TB/month after adding redundancy At the scale of Huggingface, that still amounts to a lot load of money. Significantly less than if you did the same in AWS, but still a lot That said, they do have a deal with AWS to make the data available in AWS ip space. Maybe they got some ch…

It's less than $1/TB. If you have settlement-free peering it's like fractions of a penny at scale. Cloud provider DTO pricing is literally the biggest scam on Earth.

What does DTO mean in this case? And settlement-free peering? I didn't know the price for S3 could be so cheap with AWS

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#248

Earlier quoted context omitted.

Sorry we a focusing on leather handbags there.

The leather bag designs are shipped from Paris or Milan to China for production.

It depends on the bags. I know people who work in handbags production, in France.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#249

Earlier quoted context omitted.

Indeed. My fellow software engineers keep complaining about using up all their Claude tokens within an hour... Whilst I'll be rocking DS flash for the entire day. Sure it gets a few things wrong here and there, but that's when you pull out the Claude models or whatever for those tricky tasks.

what plan are your 'fellow software engineers' using? I have a hard time even using up the Fable part of my allowance in a week of coding.

I'm on the Premium business plan and I can easily use up my entire week's allocation in two days of coding. I only use Claude for planning, too, the rest is done by Deepseek and GPT.

Claude is very spendy.

Re: DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

#250

Earlier quoted context omitted.

Personally I'm hoping for EU to step up a bit

We'll make sure there will be enough regulations to have neither the electricity to power it or the companies to build it.

Well quite a few nuclear powerplants have had to shut down recently due to droughts... So we are well on the way for that!
Post reply on HN