> ## Documentation Index
> Fetch the complete documentation index at: https://docs.somark.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> OXR is SoMark's open-source document parsing model, with a free API compatible with the SoMark API, CLI, and SDK

## What is OXR

OXR (Optical Everything Recognition) is an open-source model from the SoMark team. It turns PDFs and document images into Agent-Friendly data in Markdown or JSON, with layout coordinates included.

Pretrained from scratch, OXR uses a recursive self-improvement (RSI) loop during post-training—a first in document parsing. With just 1.3B parameters, OXR ranks #1 on olmOCR-Bench, OmniDocBench, and DocParsingBench while delivering industry-leading parsing speed.

OXR is the first to make its layout analysis module independent and pluggable. This design accelerates parsing and supports replacing the built-in layout model with a custom one.

> **[OXR on GitHub](https://github.com/SoMarkAI/OXR)**
>
> Source code, Docker deployment, and model downloads

> **[Online demo](https://free-oxr.somark.ai)**
>
> Upload a document and see the parsing result, no sign-up needed

## Relationship with SoMark

OXR's API is compatible with the SoMark API. Although OXR does not support some specialized capabilities, SoMark's ecosystem fully supports deployed OXR services.

* A self-hosted OXR and the Free OXR API both work with SoMark's official CLI and SDK tools ([somark-py and somark-js](/en/cli-sdk/index.md)). Just change the **Base URL**.
* Paths, parameter names, and response structure match the [SoMark API](/en/api-reference/index.md).

> [!WARNING]
>
> OXR only implements the core parsing capabilities. Some advanced parameters are not available, such as cross-page merging, title level recognition, `zip` output, and PDF to Office. These parameters are ignored or rejected with a parameter error. See [API alignment](/en/api-reference/oxr/alignment.md) for the full comparison.

## Free API

OXR provides a free API so you can test parsing on your own documents without registering or requesting an API key.

We'll keep the free API online until we can no longer afford to run it.

> [!WARNING]
>
> We do not recommend using the Free OXR API in production. We do not guarantee service stability, and requests are queued, so processing speed may vary slightly.

| | somark.ai |
| - | - |
| Online Demo | [https://free-oxr.somark.ai](https://free-oxr.somark.ai) |
| API URL | `https://free-oxr.somark.ai/api/v1` |
| API documentation | [https://docs.somark.ai/api-reference](https://docs.somark.ai/api-reference.md) |

> [!NOTE]
>
> somark.cn is a mirror of somark.ai. Choose the domain that works best for your region.

### Upload limits

> **File size**
>
> Up to **10MB** per file

> **Page count**
>
> Up to **20 pages** per request

Files over these limits are rejected. For larger or longer documents, use [SoMark production](/en/api-reference/index.md) (200MB, 300 pages) or self-host OXR.

### Usage

The Free OXR API requires no authentication. Set `api_key` to any value, such as `dummy`.

```bash CLI theme={null}
# Python
pip install somark
# or JS
npm install -g somark-js

somark --base-url https://free-oxr.somark.ai/api/v1 --api-key dummy \
  parse document.pdf --formats md,json --out ./parsed-output/
```

```bash cURL theme={null}
curl --fail-with-body https://free-oxr.somark.ai/api/v1/parse/sync \
  -F 'file=@document.pdf' \
  -F 'api_key=dummy' \
  -F 'output_formats=markdown' \
  -F 'output_formats=json'
```

## Self-hosted OXR

You can deploy OXR with Docker on your own GPU server. Use `http://<host>:8000/v1` as the Base URL. The 10MB / 20-page limits do not apply. See the [OXR repository](https://github.com/SoMarkAI/OXR) for deployment steps.

> **[API alignment](/en/api-reference/oxr/alignment.md)**
>
> See the parameter-by-parameter comparison between OXR and SoMark

## Star OXR on GitHub ⭐

[![Star OXR on GitHub](https://raw.githubusercontent.com/SoMarkAI/image_bed/main/OXR_README/assets/OXR_Star_on_GitHub.gif)](https://github.com/SoMarkAI/OXR)

> [!TIP]
>
> If this project helps you, please click ⭐ Star to support us. It's the best encouragement you can give the developers!

> **[Star OXR on GitHub](https://github.com/SoMarkAI/OXR)**
>
> github.com/SoMarkAI/OXR


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.