Credit Usage
How extraction requests consume credits.
Struct PDF charges credits per extraction request, with a pricing model designed to stay easy to predict.
Simple
1 credit/page• Complex2 credits/page• No parse step
How Struct PDF Differs
| Workflow | Credits |
|---|---|
| Struct PDF simple extraction | 1 credit per page |
| Struct PDF complex extraction | 2 credits per page |
| Separate parse step required | No |
Unlike parse-first workflows, Struct PDF does not require a separate parsing request before extraction. You send the file and schema once, and credits are computed on the extraction itself.
How Credit Usage Is Computed
| Schema type | Credit usage |
|---|---|
| simple schema | 1 credit per page |
| complex schema | 2 credits per page |
A schema is treated as complex when either of these is true:
- it has
8or more top-level fields - it includes an array of objects such as
items[]
Otherwise, it is billed as a simple extraction.
Examples
| Request | Credit usage |
|---|---|
| 1-page receipt, 5 top-level fields | 1 credit |
| 3-page form, 6 top-level fields | 3 credits |
| 2-page invoice, 10 top-level fields | 4 credits |
2-page receipt with items[] line items | 4 credits |
You can think of it as:
credits = page count × schema rate
Where the schema rate is 1 for simple schemas and 2 for complex schemas.
Practical Advantage
Struct PDF keeps credit usage easy to understand:
- there is no separate parse charge before extraction
- simple documents cost just
1 creditper page - more complex schemas move to
2 creditsper page, with no additional pricing tiers
For the most predictable usage, keep your schema focused on the fields you actually need and avoid adding extra top-level fields unless they provide real value.