Mulmocast is an open-source CLI tool that turns web articles, documents, papers and ideas into multi-modal content—like podcasts, slides, comics, and more—using the power of generative AIs.
Do you want to turn a web article into a comic strips? Do you wish there is a way to create a Youtube shorts video based on a scientific paper? MulmoCast is the answer.
It's a really nice idea although it's implemented in TS. I've tried partially similar in python for my pitch slides / videos, not completely automated, since I was so tired of the repetitive efforts of that pitch slide work.
I'm wondering how to place the text on the right place in images?🤔 Or can gpt-imgae-1 generate with text embedded correctly as one-go, or with inpainting?
@hiroshi_doyu Texts within images are generated by gpt-image-1, and it is the result of the phrase "comic strips" in the image prompt. It is probably possible to specify the location and text in the prompt (in the generated script), but I'd rather automate it the whole process.
Exciting to see a truly AI-native approach to presentations. Seamless integration of text, voice, and visuals opens up more dynamic and engaging communication.
Mulmocast is a powerful developer tool that lets you generate video, podcasts, slideshows, PDFs, and more—using a single, JSON-based script. With seamless integration across multiple LLMs, it turns one unified input into a wide range of rich content formats. Think of it as your all-in-one content automation engine, built for devs who want scale, speed, and creative control
InstaGraph
It's a really nice idea although it's implemented in TS. I've tried partially similar in python for my pitch slides / videos, not completely automated, since I was so tired of the repetitive efforts of that pitch slide work.
I'm wondering how to place the text on the right place in images?🤔 Or can gpt-imgae-1 generate with text embedded correctly as one-go, or with inpainting?
InstaGraph
@hiroshi_doyu Texts within images are generated by gpt-image-1, and it is the result of the phrase "comic strips" in the image prompt. It is probably possible to specify the location and text in the prompt (in the generated script), but I'd rather automate it the whole process.
Exciting to see a truly AI-native approach to presentations. Seamless integration of text, voice, and visuals opens up more dynamic and engaging communication.