Quizzer turns PDFs into quiz courses

The interesting part is not just “PDF to quiz,” but the insistence on page images instead of OCR, which should preserve layout-dependent details that text extraction often drops. It’s a strong fit for dense textbooks and manuals, though the terminal-first workflow keeps it squarely in the power-user lane.

• –Rasterizing pages before generation is a real quality move for charts, figures, sidebars, and formatting-heavy material
• –The progression from true/false to free recall gives it an actual learning loop, not just quiz spam
• –SM-2 scheduling plus adaptive difficulty pushes it closer to a study system than a one-shot generator
• –The OpenAI-compatible backend keeps it flexible for local or hosted models
• –Terminal UX will limit mainstream adoption, but it’s a sharp open-source prototype for serious self-study