The best way to scan documents in 2026 (a pragmatic ranking)

Strip away habit and marketing, and “how should I scan this?” has a short decision tree in 2026. Here it is, situation by situation.

The decision tree

• Everyday document, any device, care about privacy → browser-based on-device scanning (ToPDF): no install, no account, clean multi-page PDFs, verifiably local.
• Deep in one ecosystem, scans belong in that cloud → the ecosystem's tool (Drive scan, Notes, Lens) — least friction to that destination.
• Daily volume + searchable-text archives at work → a cloud suite (Adobe-class) or office hardware; you're buying OCR-at-scale and workflow.
• Archival photos/art → flatbed glass, borrowed if needed.

What changed to make the browser the default

1. WebAssembly matured: real computer vision and neural models now run in the tab at practical speed.
2. PWAs closed the app gap: offline, home-screen icons, share-sheet integration.
3. Privacy pressure rose: people started asking where the passport photo goes; “nowhere” became a feature worth building for.

Whatever you choose: the three constants

Good light, steady square-on framing, contrast with the background — the checklist outranks tool choice for output quality. And name your files like future-you will search for them.

Frequently asked questions

Is there one tool that covers everything?

No single tool wins every row above, and that’s fine — pick per situation. The browser option’s superpower is being available in all of them.

What about AI photo-to-document features in phone galleries?

Gallery “document” modes are improving at cleanup but typically stop short of proper multi-page PDF workflows. Useful capture aid; not the whole pipeline.