API Pricing

Pay by the page. Across every format.

One simple unit for entity recognition, redaction, and aliasing, whether you send a PDF, a Word doc, a spreadsheet, a CSV, or a raw JSON payload. No surprises from file size or format.

Estimate your cost
1 page = 1,800 characters
PDFDOCXPPTXXLSXCSVJSON

We normalize every file to the text inside it (~300 words per page), then count. You're billed for what we actually process, never for file bloat.

Billed precisely per character. The page is just a friendlier unit, and the exact character count is on every API response.

Plans

Three plans. Flat per-page rates.

Every plan includes all four endpoints (/scan, /analyze/review, /analyze/auto, /redact) plus the full policy engine, aliasing, and audit trail.

Developer
$0/mo
100 pages included
hard cap, no overage
  • All 4 endpoints, all formats
  • NER + redaction + aliasing/synthetic
  • Community support
Start free
Starter
$99/mo
1,650 pages included
then $0.06 / page
  • All 4 endpoints, all formats
  • NER + redaction + aliasing/synthetic
  • On-prem ready, audit trail
  • Email support, 99.9% uptime
Start Starter
Most popular
Growth
$799/mo
20,000 pages included
then $0.04 / page
33% volume discount vs Starter
  • Everything in Starter
  • Priority support, usage dashboard
  • Audit-log export
Start Growth
Enterprise · White-label · On-prem / Air-gapped

Committed-volume bundles or a flat annual capacity license for deployments in your own cloud or a disconnected network. Usage never leaves your environment. Includes white-label branding, custom policies, and dedicated onboarding.

Talk to our team
Estimate

See your monthly cost in seconds.

Drag to your expected volume. We pick the cheapest plan for you automatically.

pages/mo

36,000,000 characters · ≈ 6,000,000 words

Recommended plan
Growth
$799/mo
20,000 pages included
Why pages

Clear pricing, whatever you send.

Bytes lie. A 50 MB scanned PDF can hold less text than a 400 KB CSV. We price the text we actually read, so your bill tracks real work, not compression or embedded images.

We extract the text content, the cell values and field values, and count characters, then divide by 1,800. No pages required in the source file. Every format lands on the same unit.

We OCR them first, then count the recognized text like any other document. Heavy-OCR batches are metered at 2x pages to cover the extra processing, and it is always shown on your usage dashboard.

On self-serve plans, yes. /scan, /analyze, and /redact all count as pages processed. High-volume automated redaction pipelines get custom rates on Enterprise.

A typical business document runs 1 to 3 pages of text, a contract 10 to 40. Use the estimator above to size it, and once you are live your dashboard shows exact page counts per call from day one.

Priced as a flat annual capacity license instead of per-call, since usage stays entirely inside your environment. Unlimited within your deployment tier.

Ship a compliance-grade data layer in weeks.

Try the live API in the cookbook, or bring a real use case and we'll map exactly where DataGuard fits.

Published rates, subject to change. Usage is billed per character. The page is a display unit equal to 1,800 characters (~300 words), and the exact character count is returned on every API response and invoice. On-prem and air-gapped deployments are priced as a flat annual capacity license rather than per-call.

© 2026 Sidian Technologies Corp · DataGuard