GenPixy

GenPixy Blog

Guides and news for video creators

Ideas, how-tos, and updates for making explainer videos for YouTube, TikTok, Reels, and class.

GenPixy

Type a topic and get a narrated explainer video. For creators, teachers, and university students aged 18 and over.

How does GenPixy DrawGen uses AI?

  • DrawGen does not use any generative video model.
  • DrawGen does not use any generative image model.
  • AI usage is limited to just a language model. User cannot directly see its text output, and generation is strictly limited and uses strict safety guardrails. It has no image generation model and no video generation model, uses none through any other company, and does not start image or video generation for anyone. Think this service similar to a tool that converts PowerPoint Slides into animated presentations.
  • The core slideshow-generation workflow is not inherently dependent on an AI language model. We use a database of pre-made vector icons and captions then our engine animates them. These are standard stock voices, with no voice cloning, custom voices, or voice uploads. Similar functionality was possible before the ChatGPT era using conventional techniques such as deterministic text matching or rule-based logic.
  • Why this is not AI image or video generation: no model creates any image, frame, or footage, including the thumbnail. The icons exist before the customer arrives (we have a database of vector assets). The typed topic is not a prompt to an image or video model; it is the subject for the script writer. There is also no prompt builder, model picker, batch takes, spend ceiling, or bring-your-own-key option. Because DrawGen has no image or video generation model, customers cannot choose an image or video model and there is no image/video generation workflow for them to start. We do not serve AI image/video generation models.
  • It cannot create photorealistic footage, realistic depictions of people, face swaps, or deepfakes. The animation maker accepts no photos, images, video, audio, or voice samples, and there is no voice cloning.
© GenPixy