TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: Real-time image generation with SDXL Lightning

444 pointsby treesciencebotabout 1 year ago

40 comments

altryne1about 1 year ago
I&#x27;ve used this + Groq yesterday to augment (with a chrome extension) the infinite fun game from Neal Agrawal, but generate actual images and not only emojis.<p>This feels like the future, near real time image and LLM generation (using Mixtral from Groq as my prompt writer) and Fal API for read time generation!<p><a href="https:&#x2F;&#x2F;x.com&#x2F;altryne&#x2F;status&#x2F;1760561501096575401?s=20" rel="nofollow">https:&#x2F;&#x2F;x.com&#x2F;altryne&#x2F;status&#x2F;1760561501096575401?s=20</a>
评论 #39477958 未加载
评论 #39476214 未加载
评论 #39475344 未加载
评论 #39478367 未加载
评论 #39475984 未加载
minimaxirabout 1 year ago
For posterity, SDXL Lightning is open source on Hugging Face with a relatively-permissive license: <a href="https:&#x2F;&#x2F;huggingface.co&#x2F;ByteDance&#x2F;SDXL-Lightning" rel="nofollow">https:&#x2F;&#x2F;huggingface.co&#x2F;ByteDance&#x2F;SDXL-Lightning</a><p>There are a few other UIs for it, e.g. <a href="https:&#x2F;&#x2F;replicate.com&#x2F;lucataco&#x2F;sdxl-lightning-4step">https:&#x2F;&#x2F;replicate.com&#x2F;lucataco&#x2F;sdxl-lightning-4step</a>
评论 #39474926 未加载
评论 #39477290 未加载
评论 #39476869 未加载
krasinabout 1 year ago
Great speed!<p>As for the quality, I borrowed the query ([1]) that people used to test Stable Diffusion 3 and other models today: &quot;Photo of a red sphere on top of a blue cube. Behind them is a green triangle, on the right is a dog, on the left is a cat&quot;.<p>Here is what I got: <a href="https:&#x2F;&#x2F;imgur.com&#x2F;a&#x2F;XrAuqCB" rel="nofollow">https:&#x2F;&#x2F;imgur.com&#x2F;a&#x2F;XrAuqCB</a><p>To compare it with Stable Diffusion 3: <a href="https:&#x2F;&#x2F;pbs.twimg.com&#x2F;media&#x2F;GG8mm5va4AA_5PJ?format=jpg&amp;name=large" rel="nofollow">https:&#x2F;&#x2F;pbs.twimg.com&#x2F;media&#x2F;GG8mm5va4AA_5PJ?format=jpg&amp;name=...</a><p>1. <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=39467526">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=39467526</a>
评论 #39476011 未加载
评论 #39476431 未加载
评论 #39477476 未加载
refulgentisabout 1 year ago
_Really_ impressive demo but it&#x27;d be oh-so-much-more-impressive if it was smooth, right now ex. deleting a word or adding a space causes 4 inferences in quick session so it feels janky (EDIT: maybe intentional? step by step displayed?)<p>Btw this is from fal.ai, I first heard of them when they posted a Stable Cascade demo the morning it was released.<p>They&#x27;re *really* good, I *highly* recommend them for any inferencing you&#x27;re doing outside OpenAI. Been in AI for going on 3 years, and on it 24&#x2F;7 since last year.<p>Fal is the first service that sweats the details to get it to the point it runs _this_ fast in practice, not just in papers. ex. web socket connection, use short-lived JWTs to avoid having to go through an edge function to sign a request with an API key, etc.
评论 #39476263 未加载
Reubendabout 1 year ago
I love this demo. It&#x27;s easily accessible, fast, and intuitive. It&#x27;s stunning that we can get this level of quality so easily.
评论 #39488030 未加载
actionfromafarabout 1 year ago
How ... can it be so fast? And what is &quot;blob:<a href="https:&#x2F;&#x2F;blbahblah" rel="nofollow">https:&#x2F;&#x2F;blbahblah</a>&quot; image?<p>By the way the racoon is very prone to get two tails if you change the prompt a little. :)
评论 #39475418 未加载
评论 #39475287 未加载
danielecookabout 1 year ago
This is incredible. The reduction in latency has a profound effect in terms of the way I interact with this type of tool. The speed benefit is more than just more image generation. The speed here helps me easily keep the same train of thought moving along as I try different things.
thyroxabout 1 year ago
Wow this is super impressive but does somebody know a way to generate consistent characters with stable diffusion?<p>What I mean is if my first prompt is a girl talking to cat and second prompt is girl playing with that cat, I want the girl and cat to be the same in both pictures.<p>Is that possible? If so any links or tutorials will be super helpful to learn.
评论 #39477679 未加载
评论 #39479606 未加载
评论 #39478978 未加载
评论 #39479405 未加载
评论 #39482020 未加载
评论 #39478330 未加载
评论 #39491462 未加载
pugworthyabout 1 year ago
It&#x27;s curious what it does with single letters. Seems for me to often settle on a small rather detailed building. The more I repeat the letter (e.g., &quot;111&quot; vs &quot;11111111&quot; the odder the building gets. Which I can see now seems pretty sensitive to the seed.
评论 #39477680 未加载
airstrikeabout 1 year ago
Absolutely love this. Wish URLs were shareable!<p>`late 90s movie poster, 24 hour clock movie &quot;2: Electric Boogaloo&quot; dan aykroyd1`<p>turned out great
评论 #39478851 未加载
评论 #39478282 未加载
评论 #39477488 未加载
Glyptodonabout 1 year ago
Speed is impressive, but it doesn&#x27;t seem to know what a pappenheimer rapier is. Keep getting a guy with weird bladed staff.
评论 #39477061 未加载
albert_eabout 1 year ago
Idea: convert this into a side scrolling game where the background gradually and seamlessly transitions into a rendering of the words we are dealing with as we progress.<p>I am imagining the green lush landscape from early parts of the demo to slowly transform into the dry mountainous landscape from later images while new characters appear in the foreground.
tdudhhuabout 1 year ago
Wow this is fast!<p>The results also look great, but the more I see AI generated images the more I believe it is not going to eat jobs.<p>Almost none of the results are production ready. They all contain strange parts. And maybe more important: the look and feel is always the same.<p>It is amazing this works as fast as it does, but I think AI is still in it&#x27;s &#x27;hype&#x27; stage.
评论 #39478469 未加载
评论 #39478999 未加载
hocabout 1 year ago
&quot;bug with long antennas and chrome legs&quot;<p>you only end there because it&#x27;s so fast. nice. really a new explorative quality.
supermattabout 1 year ago
What is the difference between SDXL turbo (released last november) and lightning (released 2 days ago?)? I havent seen any discussion of lightning on here the last few days and hnsearch only shows a few posts with no comments.
评论 #39478451 未加载
kennyadamabout 1 year ago
Incredibly impressive how fast it is.<p>Very disturbing results are just a second away!<p><a href="https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;q5oe57udw953" rel="nofollow">https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;q5oe57udw953</a>
tiborsaasabout 1 year ago
Insanely cool :)<p>One bug though: when I increment the seed, it renders two image and it takes a bit of jumping up and down in seed numbers to get to the same image.
jansanabout 1 year ago
Very cool. How about putting the images that were created along the way into a carousel that you can swipe through, just like a flipbook?
评论 #39479091 未加载
SeanAndersonabout 1 year ago
wow! this is really fast. this is actually at the speed I want image generation to be at. it makes me want to explore and use the tool. everything else has felt like begrudgingly throwing a prompt and eventually giving up because the iteration cycle is too slow.<p>of course the quality of what is being generated is not competitive with SOTA, but this is going in a really good direction!
diego_sandovalabout 1 year ago
This is amazing. How much does it cost per image?
tobeagramabout 1 year ago
I had some fun inputting the American Psycho transcript into this. Made a little video [1]. Insane how fast this is, great work!<p>[1]: <a href="https:&#x2F;&#x2F;x.com&#x2F;TobiasWhetton&#x2F;status&#x2F;1761038697074466868" rel="nofollow">https:&#x2F;&#x2F;x.com&#x2F;TobiasWhetton&#x2F;status&#x2F;1761038697074466868</a>
ongytenesabout 1 year ago
Looking forward to it once it is refined. It still creates some strange artifacts. Within the HN message, the demo image of the baby raccoon in priest robes, one can count five paws.
monkeydustabout 1 year ago
This is nuts, in a good way. Way more fun to have fast inference on images than text.
padolseyabout 1 year ago
This is so quick! The demo being publicly consumable is powerful, though SDXL is not immune to abuse or NSFW (and possibly illegal) generations. I wonder who ends up being held accountable for such generations.
评论 #39479218 未加载
edf13about 1 year ago
Still issues with limbs when using AI generative artwork...<p><a href="https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;0y54bhvalknv" rel="nofollow">https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;0y54bhvalknv</a>
petermcneeleyabout 1 year ago
Who is gonna tell George Lucas? <a href="https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;o8d16q87qkbt" rel="nofollow">https:&#x2F;&#x2F;fastsdxl.ai&#x2F;share&#x2F;o8d16q87qkbt</a>
评论 #39483078 未加载
rcarmoabout 1 year ago
Ooooookaaaaaay.... This is faster than autocomplete. Amazing.
bsenftnerabout 1 year ago
Seems like with this kind of speed, the only reason this is not a video tool is the lack of consistency controls?
ultimooabout 1 year ago
inferencing images at the speed of typing is such a fantastic way to show off what it&#x27;s capable of. kudos!
评论 #39478058 未加载
timnetworksabout 1 year ago
That&#x27;s very fast... typing in SDXL as the prompt gives a chilling self-portrait though.
superasnabout 1 year ago
What would be the cost per month to host the same demo yourself?
doodlebuggingabout 1 year ago
&quot;A cinematic shot of a coelacanth armed with a yellow corinthian leather hand cannon&quot;<p>Needs work.
评论 #39475725 未加载
ozfiveabout 1 year ago
It&#x27;s so fast I question the quality of all other models that cost money.
a1oabout 1 year ago
Yeah it can&#x27;t do pixelart or any other thing that I tried. But yey speed.
评论 #39476043 未加载
ritzacoabout 1 year ago
This is so impressive. Is it an ad for fal.ai? Or who is paying for it?
victorbjorklundabout 1 year ago
Damn that is fast.
Workaccount2about 1 year ago
Mind blowing how fast this is
_ache_about 1 year ago
Still doesn&#x27;t pass the Axolotl test. :&#x2F; I guess it need more memory.
rcarmoabout 1 year ago
Note: doesn&#x27;t seem to work on iPad.
评论 #39487569 未加载
fareeshabout 1 year ago
Nancy Pelosi as a ninja without a mask = Nancy Pelosi as a ninja wearing a mask