Powered by Marker — the best open-source PDF parser for equations

Marker, hosted.

The scientific PDF parser ML engineers already trust. Now a one-line API. No CUDA, no 5GB model downloads, no GPU server to babysit.

Try Free on RapidAPI See benchmark ↓

2 free papers/month · No credit card · API key in 30 seconds

You already know Marker is the right tool.

You just don't want to run a GPU server to use it.

SELF-HOST MARKER
  • Download 5GB of model weights
  • Install CUDA + PyTorch + torchvision + transformers
  • Provision a GPU server and keep it warm
  • Build a queue so requests don't time out
  • Monitor, scale, patch, pay idle GPU bills
  • 2+ days of setup. $60+/mo idle cost.
PAPERPARSE
  • Click subscribe on RapidAPI
  • Copy your API key
  • POST a PDF URL. Get markdown back.
  • Same Marker quality. Same LaTeX equations.
  • Pay only for papers you parse.
  • 30 seconds to first paper. $0 idle.
COMING SOON

Head-to-head benchmark

We're running Marker, Mistral OCR, LlamaParse, Docling, and OpenDataLoader on 10 real papers across arxiv, biorxiv, and published journals. Publishing equation accuracy, citation linking, and table fidelity. No cherry-picking.

Benchmark drops next week. Follow along or subscribe to get the results.

Works on any scientific PDF

Not just arxiv. Real research comes from journals, preprint servers, pharma reports, and technical PDFs that live outside arxiv. We parse all of them.

arxiv
CS, physics, math
biorxiv / medrxiv
biology, medicine
chemrxiv
chemistry preprints
Nature / Science
published journals
PubMed Central
open access research
IEEE / ACM
conference papers
Internal research
pharma, biotech
Technical reports
any typeset PDF

Two endpoints. That's the whole API.

STEP 1 — SUBMIT
curl -X POST \
  https://rapid-api-host/parse-paper \
  -H "X-RapidAPI-Key: $KEY" \
  -F "url=https://arxiv.org/pdf/1706.03762"

# → {"call_id": "fc-01K...", "status": "queued"}
STEP 2 — POLL
curl https://rapid-api-host/parse-paper/$ID \
  -H "X-RapidAPI-Key: $KEY"

# → {
#     "status": "done",
#     "result": {
#       "title": "Attention Is All You Need",
#       "markdown": "# ...$$...$$...",
#       "char_count": 47112
#     }
#   }

Typical parse: 60–180 seconds. Async polling pattern means no 30s timeouts to fight.

What you actually get back

LaTeX equations preserved

Display math $$...$$ and inline math $...$ extracted exactly. No formula-not-decoded placeholders.

[12]

Citations linked

Inline references in the body link to entries in the References section. No manual matching.

Tables as markdown

Real markdown tables, not images, not lost. Drop straight into your vector store.

§

Section structure

Title, abstract, and headings extracted automatically. Perfect for chunking.

Async, no timeouts

Submit, get a call_id, poll. Built for the realities of parsing 50-page papers.

🎯

Built on Marker

Scored ~10.5/12 on equation extraction. Closest open-source alternative scored ~5/12.

Simple usage-based pricing

Free to start. Pay only for papers you actually parse.

Basic

$0
forever
  • ✓ 2 papers / month
  • ✓ $0.30 / paper after
  • ✓ Full equation extraction
MOST POPULAR

Pro

$9/mo
75 papers included
  • ✓ 75 papers / month
  • ✓ $0.15 / paper after
  • ✓ Email support

Ultra

$29/mo
400 papers included
  • ✓ 400 papers / month
  • ✓ $0.10 / paper after
  • ✓ Priority support

Mega

$99/mo
2,000 papers included
  • ✓ 2,000 papers / month
  • ✓ $0.10 / paper after
  • ✓ Priority support
View all plans on RapidAPI →

Ready to actually parse the equations?

Free tier. No credit card. 5 papers to test it on your real workflow.

Get Your API Key →