Per-Document Pricing: Examples & Companies

What is it

Per-Document Pricing is a billing unit where customers are charged per document processed or generated — common in AI writing, SEO, and document-intelligence tools. It is the most deliverable-shaped unit in usage pricing: the thing being counted is the thing the buyer actually wanted, a finished article or a processed file, which makes the bill legible in a way tokens and credits never are.

The AI writing and SEO cluster carries the unit most visibly. Byword is the purest case — its $99/$299/$999 tiers are 25, 80, and 300 articles a month with published per-extra-article overage. Frase sells 10 to 100 AI-optimized articles per month across its ladder; Surfer SEO bundles 120 to “unlimited” documents into seat-tiered plans; Scalenut and Writesonic package articles-per-month alongside audits and tracked answers in quota grids.

A second group uses the document as a capacity or ingestion unit rather than a deliverable: Dify caps knowledge-base documents per workspace tier (50 on Sandbox, 500 on Professional), Nomic meters documents flowing into its embedding and mapping platform, and Unbabel’s Widn.ai limits file uploads per month (5 free, 25 on the $19 Plus tier).

How it works

The standard structure is a monthly allowance with optional overage: bill = tier fee + max(0, documents − included) × per-document rate. The levers:

Worked example — the overage crossover. An agency producing 120 articles a month on Byword Standard pays $299 + 40 extra × $3.00 = $419/month — still far below the $999 Scale tier, which only breaks even around 313 articles. Allowance-plus-overage means the published tier prices are anchors, not answers; the real comparison is your volume against the marginal rate, the same arithmetic the introduction to usage-based pricing walks through for any quota system.

Worked example — document as capacity. On Dify, documents don’t bill individually at all: the $59 Professional workspace includes 500 knowledge documents as a standing capacity alongside message credits. The unit is the same noun, but it meters a stock (what you can keep indexed) rather than a flow (what you produce this month) — two different economic objects sharing one word.

Companies using this

8 in-corpus companies meter documents: the writing/SEO cluster (Byword, Frase, Surfer SEO, Scalenut, Writesonic) selling monthly article allowances, and the ingestion side (Dify, Nomic, Unbabel) counting documents uploaded, indexed, or embedded.

Patterns observed

The cluster prices the deliverable and hides the variance. One document is one unit regardless of length on Frase, Scalenut, and Writesonic alike — size-blindness traded for countability — and the vendors absorb long-document costs in the allowance price. Where overage exists it falls with tier (Byword’s $3.50→$2.50 ladder), making the tier grid a volume-discount curve; where it doesn’t, the cap is the upsell (Unbabel’s 5 free uploads, Dify’s 50-document sandbox).

The second pattern is allowance-as-discount: Scalenut runs a standing annual promotion that doubles the article quota at the same sticker price — repricing through the denominator rather than the price, the quota-system move that lets a vendor discount without ever printing a lower number. And at the top of every ladder sits a softened cap: Surfer SEO’s “unlimited*” documents on Peace of Mind convert the meter into a fair-use promise exactly where heavy users would otherwise churn to overage-free competitors.

Counterexamples & variants

Writesonic shows the unit losing its headline status: articles per month still appear in every tier, but the value metric the packaging now leads with is tracked AI answers — the company migrated its meter from words to documents to answer-coverage as the buyer’s definition of value moved, leaving documents as one quota among several. Nomic buries the unit even deeper: documents are what its platform embeds and maps, but the bill is $40 seats feeding a pooled AI-usage balance under a $1,000/month commitment — the document is the workload, not the line item. And Dify’s capacity-style cap is the structural variant worth naming: when documents meter a knowledge base rather than a content pipeline, the right comparison set is storage pricing, not article allowances.

What this means for buyers vs vendors

For buyers

Compute the effective per-document price at your volume — allowance fee divided by documents you’ll actually use, plus marginal overage beyond — rather than comparing stickers; the spread runs from ~$1.33 to ~$4.90 per article across this cohort before overage. Check rollover (usually none), check what one document includes (Frase bundles research and optimization; a bare generation elsewhere may need three more tools), and treat “unlimited” tiers as fair-use contracts to read, not ceilings to ignore.

For vendors

The document is a unit buyers trust because they can count their own output — protect that by keeping one document = one unit and putting variance absorption in the allowance price, not in surprise multipliers. Publish the overage rate and let it fall with tier (the credit-model guide covers why visible marginal pricing beats opaque packs); if your documents are stocks rather than flows, price capacity explicitly instead of dressing storage limits as usage meters. And when the buyer’s value metric moves — as it moved from articles toward answer-coverage for the SEO cluster — move the headline unit with it before a competitor does.