Drop a PDF, Word doc, or plain text. raw2slides' AI reads it, picks a design system, rewrites your content into slides, and fetches matching visuals — all in one command.
Gemini analyzes your content and selects the optimal design system — color palette, typography, image style, and layout density. No templates. Fully AI-driven.
Compresses or expands your content into exactly the slide count you specify (1–20). Applies Executive, Briefing, or Training mode automatically based on scope.
Gemini generates precise image keywords. The engine queries Unsplash → Pexels → Pixabay with automatic fallback. Images are center-cropped and embedded pixel-perfectly.
Enterprise-grade output in a single CLI command. Works with PDFs, DOCX, PPTX, and plain text — no GUI, no drag-and-drop theatre.
High-level goals, key metrics, immediate actions only.
Objectives, section dividers, knowledge checks, recaps.
Problem, framework, solution, timeline.
Full table extraction from DOCX. Shape text iteration from PPTX. Page-by-page parsing from PDF.
Gemini generates a laser-specific keyword per slide. The fetcher queries all three sources with automatic fallback and landscape-preference filtering.
Pass any PDF, DOCX, PPTX, or TXT file as --input. The parser extracts all text and table content automatically.
Gemini reads your content and returns a complete design system: style, tone, color palette, image mood, and text density.
Content is ranked, compressed, and mapped to your slide count. Images are auto-fetched per slide using Gemini-generated keywords.
A pixel-perfect 16:9 deck drops to output/final_deck.pptx. Optionally export to PDF with a single flag.
One command. Any document. A commercial-grade training deck in under 30 seconds. No design skills required.
raw2slides
No spam. No credit card. Free beta access.