Company

The short version: LeapOCR helps teams turn documents into usable data

LeapOCR is built for teams that want clean markdown, schema-first JSON, and model choices that make sense in production. This page is the fast read on what we offer and how to reach us.

Snapshot

Markdown, JSON, benchmark-backed models, clear credits, and direct access by email or scheduled call.

Outputs
Markdown + JSON

Readable review output and schema-first extraction from the same base OCR flow.

Models
standard-v2, pro-v2

Two production models built for different queue profiles, confidence needs, and cost targets.

Credit system
1 credit per step

Base OCR, customize, and bbox are priced independently per page on standard-v1 output routes.

New accounts
100 free credits

Enough to test real documents before you commit to recurring volume or top-ups.

Operating model

Build one ingestion flow, then choose the right output and model per queue

01

Run the model that matches the queue: standard-v2 for broad production volume, pro-v2 for harder or stricter workloads.

02

Use markdown when people still need to read the result and JSON when another system needs typed fields.

03

Add instructions for translation, normalization, or cleanup when the raw extraction is not enough on its own.

04

Attach bbox when review tools or UI overlays need geometry, not by default.

Support and trust

If you are evaluating for production, the fastest next step is usually one of three things: test with the free credits, send a real workflow by email, or book a call to walk through requirements.