Skip to content

WorkWingman — Study Visuals (Plain Language)

An optional feature: while you're studying for an interview, WorkWingman can generate a simple illustration to help a tricky concept click — the same idea as books that use hand-drawn diagrams to explain things (like Grokking Algorithms does for algorithms), but every picture is a brand-new AI image made just for your concept, never a copy of anyone's actual book art.

The most important thing to know: these pictures are original, not copies

WorkWingman never asks an AI to "draw it like the book does" or "recreate this specific diagram." It only ever describes the idea — "show how binary search narrows down a sorted list, with labels" — in plain, generic terms, and always tells the AI to make something completely new, not a copy of any book, comic, or artwork. The book that inspired this feature taught us that diagrams help you learn — it did not lend us its pictures, and we don't borrow them.

It costs money — and it's entirely optional

Generating an image costs a few cents, charged to your own account with Google or OpenAI, not WorkWingman's. That's why this is opt-in:

  • If you haven't connected a Gemini or OpenAI key, you'll just see a note: "Connect a Gemini or OpenAI key in Integrations to enable illustrations for study concepts." Everything else about studying still works fine — this is a bonus, not a requirement.
  • Once you connect a key (in the Integrations screen), a "Generate an illustration" box appears on the Study page. Type a concept, click generate, and a picture shows up in a little gallery you can look through, regenerate, or delete.

Which AI tools we looked at

  • Google's "Nano Banana" — really Gemini 2.5 Flash Image, Google's image-generation model, reached through Google's own developer API. Costs about 4 cents per image.
  • OpenAI's image generator — the current model in OpenAI's gpt-image family (not the older DALL·E, which OpenAI has retired). Costs somewhere between about 1 cent and 25 cents per image depending on quality/size.
  • Magnific — a tool best known for sharpening/upscaling pictures rather than creating them from scratch. So instead of a third "make a picture" option, WorkWingman offers it as an optional "Sharpen" button on a picture you've already generated — click it and Magnific cleans up and enlarges that same illustration, kept right alongside the original so you can compare. Magnific also does video, so WorkWingman adds an optional "Animate" button too: it turns one of your illustrations into a short moving clip. Both run on a monthly credit subscription (the video costs more credits than a sharpen).
  • Pixabay — we checked whether Pixabay had a way to generate AI images through an API. It doesn't (Pixabay's own AI image generator is a website-only feature, and its public API is only for searching existing stock photos) — so nothing was built here.

Every key you paste in is your own account, your own cost — WorkWingman never buys images, enhancements, or videos for you or bills you directly. The video button only appears if you've added a Magnific video key, and it's the priciest option, so it's entirely opt-in.

We also looked at another AI video tool, Revid.ai, which turns a whole script into a video with narration and captions. It's a real option but a heavier, more expensive one ($39–$199/month), and a different shape from "animate this one picture," so it's written up as a future possibility only — not built.

What actually leaves your computer

When you generate a picture: only the short concept text you typed goes out — to whichever service you picked, using your own key. Nothing else about you — not your resume, not the job posting, not your personal details. The picture comes back and is saved on your own computer.

When you Sharpen or Animate: this one's a bit different, and worth being honest about. To sharpen or animate a picture, the picture itself (the one WorkWingman already made for you) gets uploaded to your own Magnific account so it can be processed. That's the study image, not your personal information — but it is an upload, which is why the app shows a clear note about it whenever those buttons are available. It's still your key, your account, your cost.

A fair warning: AI pictures can be wrong

These are AI-generated illustrations meant to help a concept click faster — they are not a verified, textbook-accurate diagram. Treat them the way you'd treat a quick sketch a study buddy drew on a whiteboard: helpful for the "aha" moment, but double-check the actual facts before you rely on them in an interview.

Where these pictures show up

An illustration you generate attaches to whatever you were studying when you made it — a flashcard, a study-plan topic for a job. If you later export your study material into a Claude Project or NotebookLM, the illustration (and the exact description used to make it) can ride along as an extra visual aid.