Extract
Structured-extraction API that returns typed JSON matching a user-defined schema, with a citation and a verification result on every field.
What it does
- Document extraction
Sources
- Pricing
-
Per-processor rate on the pricing page; Free, Team ($400 per month) and Enterprise plans draw on the same rate card. Page quote: "Extraction — turbo / fast Structured fields, single pass. Turbo returns JSON only; fast adds per-field citations and runs on Convert 4.50"
- Description
-
Official documentation · 16 Nov 2026
Page quote: "Extract returns fields that match your schema, each one cited, verified, and scored."
- What it does
-
Official documentation · 16 Nov 2026
Page quote: "Extract returns fields that match your schema, each one cited, verified, and scored."
- Sold within
-
Official documentation · 16 Nov 2026
Sold on its own: the pricing page gives this processor its own rate per 1,000 pages. Page quote: "Extract returns fields that match your schema, each one cited, verified, and scored."
- Name
-
Official documentation · 16 Nov 2026
Page quote: "Extract returns fields that match your schema, each one cited, verified, and scored."
- URL
-
Official documentation · 16 Nov 2026
Page quote: "Extract returns fields that match your schema, each one cited, verified, and scored."
- Deployment
-
Page quote: "Deployment Hosted ✓ ✓ ✓ Run in your own VPC — — ✓ Fully air-gapped — — ✓"
- Interfaces
-
Official documentation · 16 Nov 2026
Page quote: "Full hosted API and SDK ✓ ✓ ✓"
Also from Datalab
-
Datalab API service
Datalab's hosted document-AI platform and its per-processor rate card; the Convert, Extract and Segment processors are recorded as separate products, and the platform adds Eval, which compares processing modes, and Compose, which promotes versioned pipelines.
-
Convert API service
Document parser API that converts PDFs, Office files, spreadsheets and scans into markdown, HTML, JSON or chunks with tables, equations and reading order intact in 90+ languages.
-
Segment API service
API that finds the boundaries inside a multi-document PDF and splits it into named, page-ranged documents, each tagged with a confidence level.
Something wrong here? Send a correction — quote this product id: datalab-extract.