#Off topic
1 messages · Page 1 of 1 (latest)
Let’s take this over here
So! How do you generate markdown with ChatGPT?
Also is there any way to change the dimensions of the image to be a portrait for example
@modern wyvern
@modern wyvern ??
HEY+
oups
so i just made it generate 512x384 by default. It would be possible to add a parameter to specify the aspect ratio
How to return images with markdown is basically described in the prompt
Prompts should be visually descriptive and contain some kind of style hint or artist reference. Don't make the prompt too long (e.g. a maximum of 240 characters).
Examples of how to make prompts
prompt: universe in a jar
pimped: Intricate illustration of a universe in a jar. intricately exquisitely detailed. holographic. colourful. 3d vray render, artstation
image: 
prompt: Fox with a cloak
pimped: A fox wearing a cloak. angled view, cinematic, mid-day, professional photography,8k, 50mm lens, Pixar,Alex Ross, Tim Burton
image: 
Write the title and description of a Gangsta Rap album on climate change. Include a few album cover artwork suggestions and use the Pollinations API to create the respective images.```
I’m getting this back
It’s not exactly marked down. It’s just a hyperlink
oh strange... maybe try a few times. it usually works for me
I fixed it. Thx! Had an extra space
i prefer the computerender api, I feel like it's faster and gives better results
Image creation using the Computerender API.
Prompts should be visually descriptive and contain some kind of style hint or artist reference. Don't make the prompt too long (e.g. a maximum of 240 characters).
Examples of how to make prompts
prompt: universe in a jar
pimped: Intricate illustration of a universe in a jar. intricately exquisitely detailed. holographic. colourful. 3d vray render, artstation
image: 
prompt: Fox with a cloak
pimped: A fox wearing a cloak. angled view, cinematic, mid-day, professional photography,8k, 50mm lens, Pixar
image: 
When answering, don't include the prompt and pimped, just the image, without saying "image:", and whatever other text prompt I indicate.
Write an impressionist poem about a tomato garden After the poem, attach an adequate image by monet
Thomash works for Pollination AI I believe
ohh
i wasn't aware of computerender but will check it out
yeah looks great haha. in our case the generative image url was just a test. it was kind of obvious someone would make a professional service in that style
Sadly to pass it parameters like width, number of steps, cfg etc it needs an account so it wouldnt work in chatgpt