What is UiForm?

How it works

Go further

Jupyter Notebooks

Community

Roadmap

Introduction

UiForm Docs

Convert any document type into a format suitable for LLMs

Make Files LLM-Ready

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

Models/List

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)

Processors/Submit

Reconciles multiple dictionaries to produce a single unified consensus dictionary.

Reconcile

A route mimicking the OpenAI chat completions API with additional consensus capability.
This endpoint allows for consensus-based chat completions where multiple LLM instances 
are used to generate responses and reach agreement, enhancing output reliability.
The n_consensus parameter controls how many models will be used in the consensus process.

Completions

A route mimicking the OpenAI responses API with additional consensus capability.
This endpoint allows for consensus-based completions where multiple LLM instances 
are used to generate responses and reach agreement, enhancing output reliability.
The n_consensus parameter controls how many models will be used in the consensus process.

Responses

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 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 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 from document

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

Generate

Site

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.

Get Started

UiForm Client

Examples

Introduction

What is UiForm?

How it works

Go further

Jupyter Notebooks

Community

Roadmap

Get Started

UiForm Client

Examples

​What is UiForm?

​How it works

​Go further

​Jupyter Notebooks

​Community

​Roadmap

What is UiForm?

How it works

Go further

Jupyter Notebooks

Community

Roadmap