what_you_need
Prepare the inputs before opening the apps
- PDF documents you are permitted to process
- A protected source folder
- A separate writable output folder
- A decision about whether text, links, and forms must remain interactive
step_by_step
The working process
-
1
Protect the source set
Keep the source PDFs in their original folder and choose a different output location. Compression should create delivery copies, not redefine the archive by overwriting it.
-
2
Run preflight before compression
Add the documents in Batch Compress PDF and review ready, encrypted, signed, and invalid results. Resolve malformed files separately and leave signed or encrypted sources unchanged.
-
3
Choose Flatten or Smart
Use Flatten for the strongest size reduction in most cases, especially scanned or image-heavy PDFs. Use Smart when selectable text, links, forms, annotations, and document structure need to remain intact.
-
4
Preview representative pages
Compare a text page, an image-heavy page, and any form or linked page. Use Split preview to check the Flatten trade-off or confirm that Smart preserved the structures you need.
-
5
Process into the separate folder
Run the batch and review success, unchanged, skipped, warning, and failed results. Unchanged means the candidate was not smaller and the original bytes were retained.
-
6
Open and compare the outputs
Check representative PDFs in the viewer your recipients use. Confirm page count, links, forms, searchability, and actual file sizes before deleting or moving anything.
human_checkpoint
Review before you call it finished
- Did every intended file receive a clear result?
- Are source files untouched?
- Do page count and interactive structures still work?
- Is each successful result smaller and visually acceptable?
honest_boundaries
Limits and privacy
Practical limitation
Compression results depend on the source. PDFs already optimised or dominated by incompressible content may remain unchanged. Flatten often reduces scan-heavy documents but intentionally removes interactive structure.
Privacy boundary
Discovery, preview, compression, validation, and saved outputs run locally. No document upload, account, analytics, or telemetry is required.
faq
Questions worth answering
Can the app ever make a PDF larger?
It may build a larger candidate internally, but that candidate is discarded. The saved output preserves the original bytes when no smaller valid result exists.
Should I use Flatten for everything?
Use Flatten for the strongest reduction when page-image output is acceptable. Use Smart when selectable text, links, forms, annotations, tags, or signatures matter.