PDFjet vs Unstructured
Unstructured is an open-source toolkit you run and scale yourself. PDFjet is a hosted API that returns clean, chunked, cited data in one call.
Unstructured is a favourite for RAG preprocessing: its open-source library partitions PDFs into typed elements, and a hi-res strategy uses computer vision and OCR to pull tables out with structure preserved. You can self-host it, which some teams need.
PDFjet takes the opposite trade-off — a hosted API that returns clean Markdown, CSV, JSON, editable Word or a searchable PDF in a single call, with heading-aware chunks, per-element bounding boxes and zero data retention, and no infrastructure to run or scale.
Unstructured vs PDFjet, feature by feature
| Capability | PDFjet | Unstructured |
|---|---|---|
| PDF → CSV (tables) | Partial | |
| PDF → Markdown (RAG-ready) | Partial | |
| PDF → structured JSON | ||
| PDF → editable Word (.docx) | ||
| OCR of scans / searchable PDF | Partial | |
| AI / vision-model extraction | Partial | |
| RAG chunking (heading-aware) | ||
| Per-element bbox provenance | Partial | |
| Schema-guided field extraction | ||
| Convert files → PDF | ||
| Merge / split / encrypt | ||
| Files never stored | Partial |
Comparison reflects each vendor's public documentation as of July 2026. Capabilities change — check the source links below before relying on a specific detail.
When Unstructured may fit better
- •You need to self-host / run parsing fully inside your own environment — the open-source library is free to run locally.
- •You want to compose your own preprocessing pipeline with fine-grained control over partitioning strategies and connectors.
- •You already operate the infrastructure to scale it (the OSS library is explicitly 'not designed for production').
When PDFjet is the better call
- •You want production output in one hosted call — no library to run, scale, or keep reliable.
- •You need CSV, editable Word or a searchable PDF directly, plus convert/merge/split/encrypt — all beyond Unstructured's scope.
- •You want deterministic per-element bounding boxes and heading breadcrumbs for citations, not element metadata you assemble yourself.
- •Zero data retention out of the box, without standing up and securing your own infrastructure.
Try PDFjet in one call
Point your PDF at one endpoint and pick the output — CSV, Markdown, JSON, editable Word, or a searchable PDF. No SDK required.
curl -X POST https://pdfjet.dev/extract/md \
-H "Authorization: Bearer pj_live_..." \
-F [email protected]FAQ
Is PDFjet an alternative to Unstructured?
Yes. Unstructured is an open-source/hosted preprocessing toolkit you run and scale; PDFjet is a hosted API that returns clean, chunked, cited data in one call with zero data retention. Choose Unstructured for self-hosting and pipeline control, PDFjet to skip the infrastructure.
Does Unstructured convert PDFs to CSV or Word?
Its hi-res strategy extracts tables as HTML which you can convert to CSV downstream, but direct CSV, editable Word, and file-to-PDF conversion aren't part of its documented scope. PDFjet returns those directly.
Can I self-host PDFjet like Unstructured?
PDFjet is a hosted API rather than a self-hosted library. If you must run everything in your own environment, Unstructured's open-source library fits that; if you'd rather not operate infrastructure, PDFjet handles it.
Switching from Unstructured? Start free.
100 pages/month, every feature, zero data retention. No credit card.
Get your free API key →More comparisons
Sources: docs.unstructured.io/open-source/introduction/overview·unstructured.io/pricing·unstructured.io/blog/how-to-process-pdf-in-python