Retab Docs

Introduction

Lists the currently available models, and provides basic information about each one such as the owner and availability.

Models/List

Reconciles multiple dictionaries to produce a single unified consensus dictionary with a likelihood for each key.

Reconcile

A route for converting any document into OpenAI-compatible chat messages.
You can choose between different preprocessing parameters according to your needs: modalities (text, image, native) and image settings (dpi, browser_canvas, etc..).

Create OpenAI Messages

A route for converting any document and a json schema into OpenAI-compatible responses input.
You can choose between different preprocessing parameters according to your needs: modalities (text, image, native) and image settings (dpi, browser_canvas, etc..).

Create OpenAI Inputs

A route for extracting structured data from documents using LLMs.
This endpoint processes document data and extracts information according to the provided JSON schema.
It supports various document types and can return either a complete response or streamed chunks.

Extract

Parse a document and extract text content from each page.
This endpoint processes various document types and returns structured text content
along with usage information. Supports different parsing modes and formats.

Parse

Generates a JSON Schema from scratch by inferring structure from the content of the provided example documents.

Generate

Site

Community

API Reference

Support

Create a new evaluation with documents, schema, and ground truth.
Will store any documents that don't already have a file_id.

Create

Delete

List evaluations with pagination support.

List

Update specific fields of an existing evaluation, such as name, project_id, or json_schema.
If json_schema is updated and the evaluation has iterations, the structural content of the schema (represented by schema_data_id) cannot change.

Update

List processors with pagination and filtering.

Update an existing processor configuration.

Submit documents to a processor for extraction.
This endpoint processes documents using a pre-configured processor and extracts information
according to the processor's JSON schema and settings. It supports streaming responses and
can process multiple documents in a single request.

Upload files using either:
- document=@file.pdf (single file)
- documents=@file1.pdf documents=@file2.pdf (multiple files)

API Reference

Consensus

Documents

Schema

Platform

Get