Contenu source (brut)
<div id="bsf_rt_marker"></div>
<p class="wp-block-paragraph">Working with documents can get messy quickly.</p>
<p class="wp-block-paragraph">You might have an invoice sitting in PDF form, a scanned bank statement, a contract full of text, or a research paper with tables and multiple columns.</p>
<p class="wp-block-paragraph">Getting useful information out of those documents usually means copying things manually or sending the files through several different tools.</p>
<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><strong><a href="https://taodaily.io/how-chutes-sn64-spent-the-week-making-expensive-gpus-feel-awkward/">Chutes (SN64)</a> now has a simpler option.</strong></p>
</blockquote>
<p class="wp-block-paragraph">The subnet has added <strong>Docuextract</strong>, a community-built tool that lets you send a PDF or image and get the information back as <strong>clean Markdown or structured JSON</strong>.</p>
<p class="wp-block-paragraph"><a target="_blank" rel="noreferrer noopener" href="https://chutes.ai/app/chute/vonkaiser-docuextract?tab=playground&utm_source=chatgpt.com">Try Docuextract on Chutes</a></p>
<h2 class="wp-block-heading">So, What Can You Do With It?</h2>
<p class="wp-block-paragraph">Think of Docuextract as a tool for turning documents that are difficult for software to understand into information that software can use.</p>
<p class="wp-block-paragraph">Give it a document, and it can extract the contents while preserving useful structure such as tables, headings, and other important details.</p>
<p class="wp-block-paragraph">For example, you could use it to:</p>
<ul class="wp-block-list">
<li><strong>Extract invoice information</strong> such as vendor names, dates, line items, and amounts.</li>
<li><strong>Pull information from contracts</strong> without manually going through every page.</li>
<li><strong>Process bank statements</strong> and turn their contents into structured data.</li>
<li><strong>Read scanned forms</strong> and convert them into usable digital information.</li>
<li><strong>Extract information from research papers</strong>, including lengthy multi-page documents.</li>
<li><strong>Process batches of similar documents</strong> using a reusable template.</li>
</ul>
<p class="wp-block-paragraph">This is particularly useful when you’re building an application that needs to process documents automatically.</p>
<p class="wp-block-paragraph">Instead of building a separate pipeline for OCR, document parsing, and structured extraction, you can send the document to Docuextract and get back the format your application needs.</p>
<h2 class="wp-block-heading">From a PDF to JSON</h2>
<figure class="wp-block-image size-large"><img fetchpriority="high" decoding="async" width="1024" height="576" src="https://taodaily.io/wp-content/uploads/2026/08/image-91-1024x576.png" alt="" class="wp-image-23890" srcset="https://taodaily.io/wp-content/uploads/2026/08/image-91-1024x576.png 1024w, https://taodaily.io/wp-content/uploads/2026/08/image-91-300x169.png 300w, https://taodaily.io/wp-content/uploads/2026/08/image-91-768x432.png 768w, https://taodaily.io/wp-content/uploads/2026/08/image-91-678x381.png 678w, https://taodaily.io/wp-content/uploads/2026/08/image-91.png 1920w" sizes="(max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption">How Docuextract works</figcaption></figure>
<p class="wp-block-paragraph">One of the more useful features is the ability to tell Docuextract <strong>what information you’re looking for</strong>.</p>
<p class="wp-block-paragraph">Say you have hundreds of invoices and want to extract:</p>
<p class="wp-block-paragraph"><code>invoice number → company → date → total → tax</code></p>
<p class="wp-block-paragraph">You can define those fields in a template and have Docuextract return the information as structured JSON.</p>
<p class="wp-block-paragraph">That means the output isn’t just a pool of extracted text. It’s information that can be fed directly into another application, database, or workflow.</p>
<p class="wp-block-paragraph">For developers building document-heavy applications, that’s a meaningful difference.</p>
<h2 class="wp-block-heading">It Also Handles the Messy Stuff</h2>
<p class="wp-block-paragraph">Documents aren’t always clean digital PDFs.</p>
<p class="wp-block-paragraph">They can contain scanned pages, tables, multiple columns, receipts, mathematical notation, and other layouts that make simple text extraction unreliable.</p>
<p class="wp-block-paragraph">Docuextract combines two open-weight vision-language models to handle different parts of this problem.</p>
<p class="wp-block-paragraph"><strong>LightOnOCR-2-1B</strong> handles the fast document-to-Markdown side, while <strong><a href="https://huggingface.co/numind/NuExtract3">NuExtract3</a></strong> handles structured information extraction.</p>
<figure class="wp-block-image size-large"><img decoding="async" width="1024" height="467" src="https://taodaily.io/wp-content/uploads/2026/08/image-92-1024x467.png" alt="" class="wp-image-23891" srcset="https://taodaily.io/wp-content/uploads/2026/08/image-92-1024x467.png 1024w, https://taodaily.io/wp-content/uploads/2026/08/image-92-300x137.png 300w, https://taodaily.io/wp-content/uploads/2026/08/image-92-768x351.png 768w, https://taodaily.io/wp-content/uploads/2026/08/image-92.png 1895w" sizes="(max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption"><a href="https://huggingface.co/lightonai/LightOnOCR-2-1B">LightOnOCR-2-1B on Hugging Face</a><br></figcaption></figure>
<p class="wp-block-paragraph">The technical setup is worth noting, but the important part is what it means for users: <strong>you don’t need to stitch several document-processing tools together yourself.</strong></p>
<p class="wp-block-paragraph">Docuextract puts those capabilities behind a single service.</p>
<h2 class="wp-block-heading">And Your Documents Stay Private</h2>
<p class="wp-block-paragraph">There’s another consideration when you’re sending documents to an AI system: <strong>what happens to the document after you upload it?</strong></p>
<p class="wp-block-paragraph">Docuextract runs on Chutes inside a <strong>hardware-attested Trusted Execution Environment (TEE)</strong> with end-to-end encryption.</p>
<p class="wp-block-paragraph">In practical terms, this is designed to keep the contents of documents private from the inference operator.</p>
<p class="wp-block-paragraph">That’s particularly relevant for the kinds of documents businesses are likely to process; invoices, financial statements, contracts, internal forms, and other sensitive files.</p>
<p class="wp-block-paragraph">With Chutes, you can be sure noone is peeping into your stuffs.</p>
<h2 class="wp-block-heading">The Technical Details</h2>
<p class="wp-block-paragraph">For those who want to go deeper, Docuextract combines:</p>
<ul class="wp-block-list">
<li><strong>LightOnOCR-2-1B</strong>, a 1B-parameter vision-language model for fast OCR and document-to-Markdown conversion.</li>
<li><strong>NuExtract3</strong>, a 4B-parameter vision-language model for structured extraction.</li>
<li>Support for <strong>262K-token text context</strong> through NuExtract3, allowing it to handle long documents.</li>
<li>Four REST endpoints covering OCR, extraction, template generation, and multi-page parsing.</li>
<li><strong>Apache 2.0 licensing</strong> for both underlying models.</li>
<li>Reported OCR throughput of <strong>5.71 pages per second on a single H100</strong> for LightOnOCR-2-1B.</li>
</ul>
<h2 class="wp-block-heading">Conclusion</h2>
<p class="wp-block-paragraph"><strong>Chutes now has a tool that can take the documents sitting in your inbox, drive, or applicati
