🔍 take a clooooooose look at any emoji, game character or photograph with our new insane hallucinatory upscalers
you can build these into any of your glifs via the new ENHANCE spell 😜
try via the links below:
comfyUI 🤝 glifs
you can now run comfyUI workflows inside of glifs - mix them into a prompt chain with GPT-4, add GPT-4vision to help with input images, or just let your GPU-poor friends run your comfy flows in the cloud
fresh new weekend glifs from the community
3D Clay Icon generator
Quirky Plush Animals
Cursed CCTV Footage Polaroids
Corporate Memphis Memes
links to play below!
glif of the day: DALL-E comics
this glif takes your story idea, e. g. "a squirrel pondering the meaning of life" + *any* style
..uses GPT to turn it into a short story..
..uses GPT to write 4 image prompts..
..and then stitches together 4 images, like so:
(link below)
experimenting with autonomous film formats, where an LLM takes a prompt like "dark philosophical post- apocalypse Santa Claus drama" and turns it into a full mini-movie
all media+script here is fully generated with models available on g l i f . a p p
MIDNIGHT SLEIGH:
The Glif StyleHunter browser extension is now LIVE - link below
find cool image styles on the web --> right-click-remix them with your prompts --> quickly get a new image in the same style
it's a super fun way to get inspired & generate with StableDiffusion *on the go*:
You can try out the new StableDiffusion 3 on glif right now - some incredible community creations and links below
SD3 is a major leap, combining DALLE's prompt following with the stylistic breadth and capabilities of Midjourney
...the worst it'll be!
You can now easily plug together all sorts of AI outputs into *your* design using a super simple design editor inside of - just add a CanvasBlock in the builder and design your layout.
here's how to make a O'RLY book cover generator in 1 min:
Glif dot app now supports 'prompt boosts' inside of your image spells - just use "+", or multiple "+++" to emphasise a part of your prompt, giving you more fine-grained control over your images!
Some Glif Chrome Extension updates:
1. it now keeps the aspect ratio of your input images (in Style Hunter mode, rolling this out to the other modes soon)
New spell in the glif alpha: Image2Text, a fun new block to build magical AI generators on :
Lots of fun options with this! For example, a glif that turns a single toddler drawing into a full story - the first image here is the input, snapped from a phone:
shipped an update to the Glif Browser Extension - just can now just ask it to change things about an image, and it'll do it :)
like, "make the hair blonde and curly"
tons of other stuff in there to completely remix any image web
try it with the link below:
made my first chrome extension: StyleHunter!
right click on any image as you browse, write a prompt and get a new image: exact same style -> new content
super fun on pinterest etc
no-code weekend hack, thank you GPT-4 + glif 😅
👇 leave a comment if you want to try v0.1
made a first
@heyglif
-powered GPT:
GENERATIVE TAROT
- speaks in riddles
- generates unique Tarot cards via glif dot app (using an SDXL prompt chain)
- each glif is an API endpoint, you can easily integrate them into your GPTs
play with it here:
a fun remix of the Achievement Patch generator by
@heyglif
user andresluisce, using 3 clever GPT calls to generate a fun 90s DALL-E Jesus patch based
here's "Jesus raves"
play here (and hit "remix" to see the clever prompt chain!):
@Alien_Housewife
@fabianstelzer
Thank you for flagging this - this slipped through our moderation and the user will be banned. Sorry you had to see this.
I love playing with samples on my OP-1, Digitakt etc, but kinda hate finding & organizing them
so I made an instant sample generator on glif 🎹
just prompt "deep moog synth bass" and go
v0.1 in quick action on an OP-1 featuring SamA 🔊🔊!:
You can add this FUNctionality to any of your glifs by using the glif spell, which lets you put glifs inside of your glifs ♻️
adding a library UI soon to make this super easy..🤞
we currently support a limited but growing number of comfyUI nodes and models:
🍜SDXL + RV51
🍜 IPAdapater + AnimateDiff + various Upscalers (including the workflows for insane creative enhancements)
full list here:
Run it here, and hit "remix" to see the prompts:
This glif combines Image2Text with a few GPT blocks, both to write the story based on the image contents and to determine the image prompts for each part of the narrative..
"real photography, fujifilm superia, full HD, taken on a Canon EOS R5 F1.4 ISO800 50MM, A red minimalistic rusty retro humanoid cyborg holding a cyborg retro kitten in middle of water logged streets of chittagong" by Poincare
3. in Free Form mode, add yourself into the mix: if you just prompt an image URL of a face (*make sure you're in Free Form mode!*) it will merge that face into the content of the selected reference image:
@ParsLuci2991
we're not sure which of the models this is running (there are few SD3 models). Having had access to the preview, best guess is this is a medium and not full one, but not sure.
controlnet recolor is so fun with the
@heyglif
extension. browse the web & remix/reskin/restyle any image to whatever you want - link+workflow below
here's the world's most prized pineapple made of pure blacktanium, the kind that hedgefunds put in their offices