All bundles Conversion bundle · 7 endpoints

Conversion

7 endpoints converting PDF, DOCX, XLSX, PPTX, CSV, HTML, and image OCR to clean markdown.

For RAG ingest, knowledge-base pipelines, and document-processing SaaS.

Endpoints in this bundle

Each endpoint is independently callable. Bundle membership is for discovery only — you do not need to opt in.

Method	Path	Credits	Summary
POST	/v1/convert/pdf-to-markdown	1	PDF to markdown.
POST	/v1/convert/docx-to-markdown	1	Word .docx to markdown.
POST	/v1/convert/xlsx-to-markdown	1	Excel .xlsx to markdown.
POST	/v1/convert/pptx-to-markdown	1	PowerPoint .pptx to markdown.
POST	/v1/convert/csv-to-markdown	1	CSV to markdown table.
POST	/v1/convert/html-to-markdown	1	HTML to markdown (clean).
POST	/v1/convert/ocr	1	Image OCR.

Recipe

Mixed-format knowledge-base ingest

Inspect the file's MIME type and route to the corresponding /v1/convert/* endpoint.
For scanned PDFs where pdf-to-markdown returns empty or very low text density, fall back to /v1/convert/ocr page by page.
Post-process the returned markdown into chunks of your preferred size, embed, and write to your store.

Sample code

Try a request

Pick a language. Click to expand the snippet.

curl

curl -X POST https://api.ollagraph.com/v1/convert/pdf-to-markdown \
  -H "Authorization: Bearer $OLLAGRAPH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://example.com/whitepaper.pdf"}'

python

import httpx, os

r = httpx.post(
    "https://api.ollagraph.com/v1/convert/pdf-to-markdown",
    headers={"Authorization": f"Bearer {os.environ['OLLAGRAPH_API_KEY']}"},
    json={"url": "https://example.com/whitepaper.pdf"},
    timeout=120.0,
)
print(r.json())

node

const res = await fetch("https://api.ollagraph.com/v1/convert/pdf-to-markdown", {
  method: "POST",
  headers: {
    "Authorization": `Bearer ${process.env.OLLAGRAPH_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ url: "https://example.com/whitepaper.pdf" }),
});
console.log(await res.json());

FAQ

Conversion bundle FAQ

Can I send the file body inline or only a URL?

Both. Pass `url` for us to fetch, or `file_b64` for an inline base64 body. Inline is capped at 25 MB; URL fetches at 100 MB.

Does pdf-to-markdown preserve tables?

Yes, as markdown table syntax when the layout is recoverable. Complex multi-column layouts with merged cells degrade to plain text rows.

What OCR engine backs /convert/ocr?

Tesseract 5 in our standard tier. Layout-aware OCR (Docling) is in pilot on a separate worker — contact us for access.

Conversion

Endpoints in this bundle

Mixed-format knowledge-base ingest

Try a request

Conversion bundle FAQ

Ship with the Conversion bundle.