Simple, honest pricing
Pay only for what you extract. No subscription. No per-seat fees. No floor. Cancel any time by not making another request.
Standard
Text, tables, images, math. For most pipelines.
- PDF, PPTX, scanned image input
- Byte-exact text from source layers
- Native PPTX chart data preserved
- Faithfulness check on every block
- JSON + Markdown output
- 100 MB max file size
Advanced
Diagrams as graphs, alt-text, reference-number linking.
- Everything in Standard
- Diagrams → Mermaid graph structure
- Image alt-text + embedded-text OCR
- Patent reference-number cross-linking
- Higher confidence on technical content
- Priority queue
Enterprise
Volume pricing, dedicated infra, SLAs, security review.
- Volume discounts from 100k pages/mo
- Bring Your Own Storage (BYOS) S3 + R2
- 99.9% uptime SLA
- SSO, audit logs, role-based access
- Dedicated support + Slack channel
- Custom data-residency
All prices in USD. Billed monthly via Dodo Payments. No setup fees.
Frequently asked
How do you charge?
Per page extracted. Standard at $0.005 and Advanced at $0.015. No subscription, no minimum, no per-seat fee. Invoices monthly via Dodo Payments.
What counts as a page?
One PDF page or one PPTX slide. We never charge for pages we couldn't process. If extraction fails on a page, you don't pay for that page.
Is there a free tier?
Yes. Every account starts with 200 lifetime free pages — no monthly reset, no expiry. Once those are used up, top up your wallet from any amount starting at $5. The playground is also free and unmetered for quick experiments.
How does billing work?
Prepaid wallet, like OpenAI or Anthropic. Add funds via card; every extracted page debits the balance at the per-page rate. Failed and cancelled jobs are never billed. Cached re-extractions are never billed. You can refill anytime from the Billing page.
What's the difference between Standard and Advanced?
Standard returns text, tables, images, and math — all the high-quality output without LLM enrichment. Advanced adds diagram graph extraction (Mermaid output), image alt-text, embedded-text OCR inside figures, and reference-number cross-linking. Use Advanced for patents, research papers, and technical decks.
Do you store my documents?
By default we retain source files for 24 hours so failed extractions can be retried, then auto-purge. You can set retention to 0 (delete on completion) per-request. Or use BYOS to keep files in your own S3/R2 bucket — we never see them.
How does this compare to LlamaParse Premium?
Standard is ~60× cheaper. Advanced is ~20× cheaper. Output quality matches or beats on patents and PPTX charts based on our benchmarks. See the LlamaParse comparison page for side-by-side examples.