DeepSeek V4 Flash 0731

Post by mshachkov » Sat, Aug 08, 2026, 12:10 AM UTC

namr2000 wrote: post #213

ak_t wrote: post #164

Note this is the 07/31 release of DSv4 flash and not the "preview" that they put out a couple months or so ago. I've been running this model locally for a week, and the preview version before that. This updated one feels like a whole tier up. It's very capable for debugging and analyzing documents/data I upload. The killer feature, IMO, is the speed. On 2x RTX Pro 6000 Blackwell, its ~8k tok/s prefill and ~250 tok/s…

What runtime are you using with the 2x RTX Pro 6000 Blackwell machine? I have the same setup and tried DSv4 Flash on vLLM and ran into a ton of kernel bugs that don't seem to have been fixed yet.

Post by w4yai » Sat, Aug 08, 2026, 12:12 AM UTC

thehamkercat wrote: post #36

542458 wrote: post #4

Kimi K3 was an interesting model only a month ago, and now we're looking at the same performance for 1/20th of the price. Wild how fast this is advancing.

And now nobody seems interested in it because the price hasn't gone down it's still $3/$15 for all providers on openrouter because of some Kimi license https://openrouter.ai/moonshotai/kimi-k3#providers

Post by est » Sat, Aug 08, 2026, 12:25 AM UTC

FernandoTN wrote: post #194

abixb wrote: post #162

Earlier quoted context omitted.

If what you're saying is true and accurate, then US-based AI labs are in big trouble. The only saving grace might be some sort of a 'national security' proclamation banning the use of state-of-the-art Chinese (and non-US) models across US federal and state governments and large enterprises (especially ones with federal government contracts), but even still, US AI labs will probably lose out massively on international…

I think you're overlooking the fact that for long-horizon tasks, even small errors compound over time and can lead to catastrophic outcomes. For simple queries, we have reached the threshold since the beginning of the year, and models are good enough from every provider to make a meaningful difference between one another. (ChatGPT, Claude, Gemini, Grok, MuseSpark, Kimi, DeepSeek, GLM...) The real unlock will be, and…

Post by AussieWog93 » Sat, Aug 08, 2026, 12:35 AM UTC

Cookingboy wrote: post #250

FernandoTN wrote: post #194

Earlier quoted context omitted.

I think you're overlooking the fact that for long-horizon tasks, even small errors compound over time and can lead to catastrophic outcomes. For simple queries, we have reached the threshold since the beginning of the year, and models are good enough from every provider to make a meaningful difference between one another. (ChatGPT, Claude, Gemini, Grok, MuseSpark, Kimi, DeepSeek, GLM...) The real unlock will be, and…

That doesn’t make sense. It’s not like SOTA models are error free, yet we still use them. You use Fable 5 right? If that’s good enough for you now, why wouldn’t a Chinese model that’s as good as Fable 5 but at 10% the cost be good enough in 6 months?

Post by rcpt » Sat, Aug 08, 2026, 12:47 AM UTC

anal_reactor wrote: post #222

abixb wrote: post #162

Earlier quoted context omitted.

If what you're saying is true and accurate, then US-based AI labs are in big trouble. The only saving grace might be some sort of a 'national security' proclamation banning the use of state-of-the-art Chinese (and non-US) models across US federal and state governments and large enterprises (especially ones with federal government contracts), but even still, US AI labs will probably lose out massively on international…

It's been true for almost every business. "Cheap and good enough" usually trumps "excellent but expensive". Ikea, McDonald's, Ryanair, AliExpress, Aldi - these brands prove that catering to poor people is more profitable than catering to rich people simply because there are so many poor people that their collective spending power outweights the one of rich people.