Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Deepseek vs Fable 5.2 Both were given the same prompt, only difference was harness. Both came out great! If I had to guess Fable 5.2 is trillions of parameters larger than Deepseek, and MUCH more expensive. With all of that said Deepseek is king here.

36,215 Aufrufe • vor 8 Tagen •via X (Twitter)

38 Kommentare

Profilbild von TechMD
TechMDvor 8 Tagen

Deepseek is doing well

Profilbild von Loktar 🇺🇸
Loktar 🇺🇸vor 8 Tagen

Yeah I have another example coming a bit later, Deepseek can't be counted out yet!

Profilbild von Rez
Rezvor 8 Tagen

Is it wrong to say I'm MUCH MUCH more hyped for qwen 4? Like Fable is great, but local models... we need more.

Profilbild von Loktar 🇺🇸
Loktar 🇺🇸vor 8 Tagen

No I'm also hyped for Qwen 4! I want to test it more but I'm trying to save tokens lmao

Profilbild von Rez
Rezvor 8 Tagen

I just want to run qwen 4 27B and qwen 4 flash next locally :(

Profilbild von X_1024
X_1024vor 8 Tagen

我很震惊,为什么你发布的项目效果都如此好,就像官方宣传自己模型效果时候的发布片。美术有用到模型生成或者图片生成模型吗?

Profilbild von Loktar 🇺🇸
Loktar 🇺🇸vor 8 Tagen

haha appreciate that! I don't use any reference images or anything, it's a lot of tweaking I use other quick models (mostly Deepseek now) to sketch stuff out basically and dial in a prompt then have other models test it. For everything I post theres at least 5-10 failures lol.

Profilbild von XOdJE
XOdJEvor 8 Tagen

@yngdl1789479 this was over API?

Profilbild von Loktar 🇺🇸
Loktar 🇺🇸vor 8 Tagen

@yngdl1789479 Yea

Profilbild von Elshayib
Elshayibvor 7 Tagen

It's like what ? 100x cheaper also

Profilbild von accelerate27
accelerate27vor 8 Tagen

That's a pretty well made ship by deepseek!

Profilbild von Hyper Bot Super Intelligent
Hyper Bot Super Intelligentvor 8 Tagen

the cheap one winning is the best part of this whole story

Profilbild von fengran5
fengran5vor 8 Tagen

fable water is good,Deepseek architecture and atomosphere is good

Profilbild von DeepSleep
DeepSleepvor 7 Tagen

DeepSeek Flash:现在都轮到我和最新的fable系列对打了吗🤡🤡🤡

Profilbild von Wayne Lowry
Wayne Lowryvor 8 Tagen

I'm pretty sure I played this game before

Profilbild von Fei-xiong Luo
Fei-xiong Luovor 8 Tagen

DeepSeek is amazing. I find PI + DeepSeek works well in CLI environment.

Profilbild von John Rood
John Roodvor 8 Tagen

same prompt but two different harnesses means two variables changed, so deepseek winning might just be the deepseek harness handling this task better. i run a fleet daily and the same model in two harnesses swings wider than two models in one harness. retries and error reading are doing most of the work.

Profilbild von EDDY VU
EDDY VUvor 8 Tagen

Totally on point, tackling this specific friction point makes a huge difference in practice.

Profilbild von 多云
多云vor 8 Tagen

永远可以信赖Deepseek

Profilbild von Aurora
Auroravor 7 Tagen

Fable5.2 这个水面很真实,就跟真的一样了

Profilbild von Ivan Fioravanti
Ivan Fioravantivor 7 Tagen

DeepSeek wins!

Profilbild von samuel Peñaloza
samuel Peñalozavor 8 Tagen

Share the prompt plase

Profilbild von linn yugio
linn yugiovor 7 Tagen

哇哦,你一定写了一个非常优秀的提示词。

Profilbild von DEATHLESS
DEATHLESSvor 7 Tagen

Fable is cooking ☠️

Profilbild von Joker
Jokervor 8 Tagen

deepseek的表现也不差啊

Profilbild von intelligence 🇦🇺
intelligence 🇦🇺vor 7 Tagen

Prompt

Profilbild von Dylan Lewis
Dylan Lewisvor 8 Tagen

harness being the only difference is the interesting bit did you log tool errors? that's usually where the gap shows up

Profilbild von Kuruś
Kuruśvor 7 Tagen

Same prompt, different harness. You compared harnesses, not models.

Profilbild von orlie
orlievor 8 Tagen

is it a full three.js implementation or is there more?

Profilbild von libiu(回官)
libiu(回官)vor 7 Tagen

同样的图,DeepSeek没输

Profilbild von Dingii.i
Dingii.ivor 7 Tagen

This is sick man

Profilbild von hixonchan
hixonchanvor 7 Tagen

prompt?

Profilbild von enpi
enpivor 7 Tagen

The anthropic models are still based on mythos architecture huge dense models. Dense model sure keep reasoning accurate but at humongous infra costs and power consumption. So if looking at the reasoning/energy consumption ratio (and so cost) chinese models definitely are on top.

Profilbild von Uğur inanç
Uğur inançvor 7 Tagen

Yeah Fable 5.2 is better but 200X better? Because it would cost that many times in this test LOL

Profilbild von Sharon Daniel ©️🎭
Sharon Daniel ©️🎭vor 7 Tagen

I think we are getting back to the point of the models being only as good as the prompts . Always better prompts will gov eyou better builds

Profilbild von Fries
Friesvor 7 Tagen

Look what I made! Look what I made!

Profilbild von Asmir
Asmirvor 8 Tagen

run it again with the setups swapped, deepseek in fable's and fable in deepseek's

Profilbild von AIHubmix
AIHubmixvor 7 Tagen

fable win in all ways

Ähnliche Videos