It’s April 2023 and currently, an AI craze is going strong. 2 weeks in AI feel like 6 months in AI in any previous year. It is hard to keep up with all the models and apps getting released on a daily basis, both commercially and open-source.
Now I wonder if there is an AI engine, or API to AI engine that can create vector graphics from text input?
I know that GPT-3.5 can create raw SVG output if you tell it to do so. However this may create a lot of output tokens that you have to pay for. High resolution paths defined by thousands of coordinate pairs can become expensive for any image that is more complex than, say a simple logo. Also you don’t get a preview image and therefore it is tedious to work with. (I have to paste SVG into an SVG editor to get a preview).
I’ve experimented with OpenAI’s GPT-3.5.
I like to work with SVG output, because it can be styled and animated with HTML and web programming skills.
 that can create vector graphics (SVG) from text?){kind=link}