Hey there, and welcome to PicFinder! PicFinder is the fastest and most user friendly image generation tool, and we're here to help you get the most of it from the start.
Start using PicFinder
Write any prompt in the PicFinder search bar to instantly start infinite image generation. You can select different AI models or image sizes & formats to obtain the perfect image in seconds. This quick, 1 minute demo introduces the basics:
How to create your prompts
Crafting the right prompt will help you get much more from each generation. Here are the essentials things you should know.
Describe your intended result clearly
Always use 5+ word prompts rather than single word prompts and follow this structure:
Subject - the character or object you are trying to portray
Background - elements surrounding the character or the context in which they are placed
Type - the overall look of the image, for instance photography, 3D, painting, illustration
Style - will usually be connected to the type but more precise, for instance anime or manga are styles of illustration, impressionism is a style of painting, etc.
You can change the order of these elements based on their importance, but it's always useful to have them all mentioned if possible. Here's a comparison of results.
Results for a single-word prompt: coffee. The images are fine and could even be used, but we can do better!
Instead, here are results for a multi-word prompt: a cup of coffee on a table, product photography, seen from the side, studio lighting, high detail, soft focus. The images now look more consistent but there's also a good amount of variety. I could use these on my coffee-lover blog already.
The order of words in your prompt matters
The closer to the beginning of the prompt a word is placed, the more it is prioritized in the generation of the image.
Here the words 3d render are at the start of the prompt and the results have a distinct 3D look.
In this example everything else about the prompt is the same, but the words 3d render are moved at the end of the prompt. The results are therefore more realistic, and while there is a sharp 3D flair, the result is more a mix of 3D and photography.
Explore with models and formats
Use the Settings panel to try out different models and image formats.
Models
The default model currently used for generation is AbsoluteReality - a general purpose model that favours a realistic or photographic style. But there are many other models to choose from. Some also realistic, others more geared towards fantasy or anime art. SDXL is a premium model that creates stunning results in just about any style. Try them all out to find your favourite!
Formats
The default image format is square - 512 x 512, or 1024 x 1024 for the SDXL model which makes larger images. But you can also create portrait or landscape images with different proportions. Image generation results also change when you use different image formats.
Add some quality keywords
There are some keywords you've probably seen used a lot in prompts. That's because they work. Here's a list of keywords you can add to your image generation, to improve the quality:
high detail or high resolution - increases details in images
soft focus or sharp focus - creates either a soft blurry effect, or a more crisp look
artstation - will make just about any image look more artistic and polished
volumetric lighting or studio lighting - will create a more professional lighting effect
If you're looking for photorealistic results, there are a few keywords specific to photography.
portrait and/or close-up - for photorealistic faces
award winning photograph - for more professional or dramatic looking photos
studio photography - for either product photography or more artistic portraits
analog photography - will create a more vintage, film-photography effect
You can combine all these keywords in images, to get great results. Here's a quick example with almost all words used:
🪄 You can do it!
PicFinder is simple but incredibly flexible and powerful. Also, our image generation is instantaneous, so if you have an idea - just try it out. Learn by experimenting with prompts, keywords, formats, and models. It only takes seconds, and the results will be stunning. You'll be an AI image generation expert in no time!