HyperAPI / index.html
nchhaya's picture
Remove TypeScript SDK mention
da5caa5 verified
Raw History Blame Contribute Delete
20.3 kB
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>HyperAPI by Hyperbots: document intelligence APIs for enterprise documents</title>
<meta name="description" content="One API to parse, extract, classify, split and redact any enterprise document. Schema-validated JSON with per-field confidence, an MCP server for agents on request, and a free tier.">
<style>
:root{--ink:#0f172a;--muted:#475569;--line:#e2e8f0;--soft:#f8fafc;--brand:#2563eb;--brand2:#0ea5e9;--code:#0b1220;--codeink:#e2e8f0;--ok:#0f766e}
*{box-sizing:border-box}
html,body{margin:0;background:#fff;color:var(--ink);font:16px/1.6 -apple-system,BlinkMacSystemFont,"Segoe UI",Inter,Roboto,Helvetica,Arial,sans-serif}
a{color:var(--brand);text-decoration:none}a:hover{text-decoration:underline}
.wrap{max-width:1080px;margin:0 auto;padding:0 24px}
header.hero{background:linear-gradient(135deg,#0b1220 0%,#13254a 55%,#0c3b6e 100%);color:#fff;padding:60px 0 52px}
.badge{display:inline-block;font-size:12px;letter-spacing:.06em;text-transform:uppercase;background:rgba(255,255,255,.12);border:1px solid rgba(255,255,255,.25);padding:4px 10px;border-radius:999px}
h1{font-size:44px;line-height:1.1;margin:18px 0 12px;letter-spacing:-.02em;max-width:900px}
.lede{font-size:19px;color:#cbd5e1;max-width:760px;margin:0 0 26px}
.cta{display:flex;gap:12px;flex-wrap:wrap}
.btn{display:inline-block;padding:11px 18px;border-radius:10px;font-weight:600;font-size:15px}
.btn.primary{background:#fff;color:#0b1220}.btn.ghost{border:1px solid rgba(255,255,255,.4);color:#fff}
.btn:hover{text-decoration:none;opacity:.92}
.facts{display:grid;grid-template-columns:repeat(auto-fit,minmax(150px,1fr));gap:18px;margin-top:34px;color:#cbd5e1;font-size:13.5px}
.facts b{display:block;color:#fff;font-size:24px;letter-spacing:-.01em}
section{padding:54px 0;border-bottom:1px solid var(--line)}
section.alt{background:var(--soft)}
h2{font-size:28px;margin:0 0 8px;letter-spacing:-.01em}
.sub{color:var(--muted);margin:0 0 26px;max-width:820px}
.grid{display:grid;grid-template-columns:repeat(auto-fit,minmax(200px,1fr));gap:16px}
.card{border:1px solid var(--line);border-radius:14px;padding:18px;background:#fff}
.card h3{margin:0 0 6px;font-size:17px}
.card code{font-size:13px;color:var(--brand)}
.card p{margin:8px 0 0;color:var(--muted);font-size:14.5px}
.meta{font-size:13px;color:var(--ok);font-weight:600;margin-top:10px}
.chips{display:flex;flex-wrap:wrap;gap:8px}
.chip{border:1px solid var(--line);background:#fff;border-radius:999px;padding:5px 12px;font-size:13.5px;color:#334155}
pre{background:var(--code);color:var(--codeink);padding:18px 20px;border-radius:12px;overflow:auto;font:13.5px/1.55 ui-monospace,SFMono-Regular,Menlo,Consolas,monospace;margin:0 0 18px}
code{font-family:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace}
.c{color:#94a3b8}.k{color:#93c5fd}.s{color:#86efac}
table{width:100%;border-collapse:collapse;font-size:15px;background:#fff;border:1px solid var(--line)}
th,td{text-align:left;padding:10px 14px;border-bottom:1px solid var(--line);vertical-align:top}
th{background:var(--soft);font-weight:600;font-size:13px;color:var(--muted);text-transform:uppercase;letter-spacing:.04em}
tr:last-child td{border-bottom:0}
.note{border-left:3px solid var(--brand2);background:#f0f9ff;padding:12px 16px;border-radius:0 10px 10px 0;color:#0c4a6e;font-size:14.5px;margin:0 0 18px}
.two{display:grid;grid-template-columns:1fr 1fr;gap:28px}
.flow{display:grid;grid-template-columns:repeat(auto-fit,minmax(150px,1fr));gap:10px;margin:18px 0}
.step{border:1px solid var(--line);border-radius:12px;padding:12px;background:#fff;font-size:14px}
.step b{display:block;font-size:12px;text-transform:uppercase;letter-spacing:.05em;color:var(--brand)}
.step.agent b{color:#7c3aed}
@media(max-width:780px){.two{grid-template-columns:1fr}h1{font-size:33px}}
ul.clean{padding-left:18px;margin:0}ul.clean li{margin:0 0 8px}
.small{font-size:13px;color:var(--muted)}
footer{padding:36px 0 48px;color:var(--muted);font-size:14px}
.links{display:flex;gap:18px;flex-wrap:wrap}
</style>
</head>
<body>
<header class="hero">
<div class="wrap">
<span class="badge">HyperAPI · by Hyperbots</span>
<h1>Document intelligence APIs for enterprise documents</h1>
<p class="lede">One API to <b>parse, extract, classify, split and redact</b> any enterprise document: invoices and statements, POs and bills of lading, contracts and NDAs, KYC packs, HR files, claims, lab reports and email. It returns schema-validated JSON with a confidence score and a source box for every field. You call it from your own applications, and an MCP server for AI agents is available on request.</p>
<div class="cta">
<a class="btn primary" href="https://apis.hyperbots.com/dashboard/register" target="_blank" rel="noopener">Get a free API key</a>
<a class="btn ghost" href="https://apis.hyperbots.com/sandbox" target="_blank" rel="noopener">Try the sandbox (no sign-up)</a>
<a class="btn ghost" href="https://apis.hyperbots.com/docs" target="_blank" rel="noopener">Read the docs</a>
</div>
<div class="facts">
<div><b>99.8%</b>extraction accuracy (reported)</div>
<div><b>&lt; 2 s</b>average latency</div>
<div><b>99.9%</b>uptime SLA</div>
<div><b>35M+</b>real documents in training</div>
<div><b>100+</b>languages</div>
</div>
</div>
</header>
<section>
<div class="wrap">
<h2>Five operations, one endpoint</h2>
<p class="sub">Upload a file once and get a <code>document_key</code>, valid for 24 hours. Then run any operation against it, in any order. Each operation reads the page, interprets layout and content, and returns structured output that's ready for the next system.</p>
<div class="grid">
<div class="card"><h3>Parse</h3><code>POST /v1/parse</code><p>Layout-aware text and Markdown with reading order, tables, labeled regions and word boxes with fonts. Handles scans, photos, multi-column layouts and low-quality PDFs.</p><div class="meta">~1 s · from $0.001 / page</div></div>
<div class="card"><h3>Extract</h3><code>POST /v1/extract</code><p>Typed fields and any number of line items. Use the built-in financial format or send a JSON schema for any document type. The advanced extractor detects the type automatically.</p><div class="meta">~2 s · $0.05 / page</div></div>
<div class="card"><h3>Classify</h3><code>POST /v1/classify</code><p>Document type with a confidence score and ranked candidates. 67 types out of the box, flags types outside the set, and supports custom classes.</p><div class="meta">~0.8 s · $0.05 / page</div></div>
<div class="card"><h3>Split</h3><code>POST /v1/split</code><p>Finds where each document in a bundle starts and ends, from content rather than page counts. Trained on more than 1M page streams.</p><div class="meta">~0.5 s · $0.05 / page</div></div>
<div class="card"><h3>Redact</h3><code>POST /v1/redact</code><p>Masks names, account numbers, IBANs, IDs and other PII with black boxes or synthetic values. Run it first to keep PII out of LLM prompts.</p><div class="meta">~3 s · $0.05 / page</div></div>
</div>
<p class="small" style="margin-top:16px">Included with every call: per-field confidence and source boxes, document-level checks such as line items adding up to the total, <code>request_id</code> / <code>model_used</code> / duration / page count for audit, and sync, async (<code>X-Async</code>) or Batch API with webhooks. Cached re-runs within 24 hours are not billed.</p>
</div>
</section>
<section class="alt" id="coverage">
<div class="wrap">
<h2>What it reads</h2>
<p class="sub">No per-format templates. HyperAPI uses a 100-type reference taxonomy in 15 categories for classification and routing.</p>
<div class="chips">
<span class="chip">Payables &amp; invoices</span><span class="chip">Statements &amp; reconciliation</span><span class="chip">Credits &amp; adjustments</span><span class="chip">Payments &amp; remittances</span><span class="chip">Collections</span><span class="chip">Procurement &amp; ordering</span><span class="chip">Shipping &amp; logistics</span><span class="chip">HR &amp; employee documents</span><span class="chip">Onboarding, KYC &amp; compliance</span><span class="chip">Financial statements</span><span class="chip">Credit, lending &amp; risk</span><span class="chip">Clinical, lab &amp; healthcare</span><span class="chip">Email, chat &amp; meeting notes</span><span class="chip">Legal &amp; HR contracts</span><span class="chip">Operations, quality &amp; facilities</span>
</div>
<div class="two" style="margin-top:28px">
<div>
<h3>Formats</h3>
<ul class="clean">
<li><b>PDF:</b> native, scanned, mixed, password-protected (with credentials), PDF forms, several hundred pages</li>
<li><b>Images:</b> PNG, JPEG, TIFF (multi-page), WebP, HEIC, BMP and more</li>
<li><b>Office:</b> DOCX, XLSX (workbooks parsed natively, sheet by sheet), PPTX, ODT, RTF</li>
<li><b>Email:</b> EML and MSG, including body blocks and attachments</li>
</ul>
</div>
<div>
<h3>Hard pages</h3>
<ul class="clean">
<li>Skewed, rotated, shadowed and low-resolution scans (deskewed and enhanced first)</li>
<li>Stamps, seals, signatures and handwritten entries, with the same confidence scoring as print</li>
<li>Dense multi-column layouts and tables with hundreds of rows, including tables that span pages</li>
<li>Mixed packets: a cover letter, a form, three invoices, a cheque and a statement in one file</li>
</ul>
</div>
</div>
</div>
</section>
<section id="quickstart">
<div class="wrap">
<h2>Quickstart</h2>
<p class="sub"><a href="https://apis.hyperbots.com/dashboard/register" target="_blank" rel="noopener">Get a free key</a> (30 pages a month, no credit card), set <code>HYPERAPI_KEY</code>, and follow the <a href="https://apis.hyperbots.com/docs/quickstart" target="_blank" rel="noopener">quickstart</a> to use the Python SDK or the REST API. An MCP server for agents is available on request.</p>
<pre><span class="k">from</span> hyperapi <span class="k">import</span> HyperAPIClient
<span class="k">with</span> HyperAPIClient() <span class="k">as</span> client:
doc = client.process(<span class="s">"packet.pdf"</span>) <span class="c"># parse + extract in one call</span>
kind = client.classify(<span class="s">"packet.pdf"</span>) <span class="c"># label + confidence</span>
parts = client.split(<span class="s">"packet.pdf"</span>) <span class="c"># start/end page per document</span>
safe = client.redact(<span class="s">"packet.pdf"</span>, mode=<span class="s">"deidentify"</span>)</pre>
<pre><span class="c"># any document, your schema (REST)</span>
curl -X POST <span class="s">"https://apis.hyperbots.com/v1/extract?category=non_financial"</span> \
-H <span class="s">"X-API-Key: $HYPERAPI_KEY"</span> \
-F <span class="s">"document_key=550e8400-e29b-41d4-a716-446655440000"</span> \
-F <span class="s">'schema={"policy_number":null,"claimant_name":null,"date_of_loss":null,"amount_claimed":null}'</span></pre>
</div>
</section>
<section class="alt" id="agents">
<div class="wrap">
<h2>Built for agents</h2>
<p class="sub">Coding agents can follow a goal and write code, but they don't read unstructured documents reliably. HyperAPI's five operations can be exposed over <b>MCP</b> (available on request from <a href="mailto:hyperapi-support@hyperbots.com">hyperapi-support@hyperbots.com</a>), so Claude Code, Cursor, GitHub Copilot, Codex, OpenCode or any MCP client can build the pipeline a goal needs. <b>Agent Builder (beta)</b> picks the operations, runs them with guardrails and returns one structured result with a citation for every field.</p>
<p><i>"Reconcile this month's vendor invoices against our open POs, flag anything over $500, and mask the bank details before it moves on."</i></p>
<div class="flow">
<div class="step"><b>classify</b>separate invoices from everything else</div>
<div class="step"><b>split</b>break the batch into single documents</div>
<div class="step"><b>extract</b>vendor, totals and line items</div>
<div class="step agent"><b>agent</b>match to open POs</div>
<div class="step agent"><b>agent</b>flag totals over $500</div>
<div class="step"><b>redact</b>mask bank details</div>
</div>
<p class="small">Other workflows built from the same five operations include AP reconciliation, contract intake, statement digitisation, supplier onboarding and mixed-inbox triage.</p>
</div>
</section>
<section id="datasets">
<div class="wrap">
<h2>Open synthetic datasets</h2>
<p class="sub">Ten fully synthetic, CC BY 4.0 benchmarks, labelled in HyperAPI's output formats. Use them to evaluate HyperAPI or any document AI pipeline without sharing a single real document.</p>
<div class="grid">
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-enterprise-doc-extraction" target="_blank" rel="noopener">Enterprise extraction</a></h3><p>31 document types, 372 docs, gold fields plus the HyperAPI schema for each.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-doc-parsing" target="_blank" rel="noopener">Parsing</a></h3><p>100 pages with layout blocks, reading order, tables and Markdown.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-doc-classification" target="_blank" rel="noopener">Classification</a></h3><p>248 pages across 31 taxonomy classes, some stamped or degraded.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-doc-splitting" target="_blank" rel="noopener">Splitting</a></h3><p>200 multi-document bundles with the correct boundaries.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-pii-redaction" target="_blank" rel="noopener">PII redaction</a></h3><p>100 pages with PII spans and boxes, plus redacted and de-identified text.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-handwritten-forms" target="_blank" rel="noopener">Handwritten forms</a></h3><p>100 forms with handwritten fields, ticks, signatures and notes.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-stamped-documents" target="_blank" rel="noopener">Stamped documents</a></h3><p>100 documents with stamps and seals over the text.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-bad-scans" target="_blank" rel="noopener">Bad scans</a></h3><p>120 pages at 3 severities, with skew, blur, noise and compression.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-dense-documents" target="_blank" rel="noopener">Dense documents</a></h3><p>60 high-resolution pages: directories, legal terms, long tables.</p></div>
<div class="card"><h3><a href="https://huggingface.co/datasets/hyperbots-inc/synthetic-invoice-extraction" target="_blank" rel="noopener">Invoice extraction</a></h3><p>200 invoices with line items and a ready-made HyperAPI scorer.</p></div>
</div>
</div>
</section>
<section class="alt" id="production">
<div class="wrap">
<h2>In production</h2>
<p class="sub">Client names are withheld. All figures are as reported by Hyperbots.</p>
<div class="grid">
<div class="card"><h3>1M+ invoices a year</h3><p>A US marketing-services enterprise with 21 entities on SAP and NetSuite. An Invoice Processing Agent built on classify, split and extract reports 99%+ extraction accuracy, 80% less manual AP work and 40% lower AP cost.</p></div>
<div class="card"><h3>$10B manufacturer</h3><p>More than 300K invoices a year posted into SAP ECC, with low-confidence fields routed to review and 40% fewer outsourced (KPO) staff.</p></div>
<div class="card"><h3>1M pages in 10 days</h3><p>An aged, dense, multi-column directory archive parsed with column boundaries and reading order intact, then extracted into structured records.</p></div>
<div class="card"><h3>One parse, five formats</h3><p>A document reconstruction pilot across vector PDF, scanned PDF, XPS, Outlook email and Excel, with 97% median text recall on the client's own files.</p></div>
</div>
</div>
</section>
<section id="pricing">
<div class="wrap">
<h2>Pricing</h2>
<p class="sub">Start free. On paid plans, usage beyond the included pages is charged to prepaid credits at per-page rates.</p>
<table>
<tr><th>Plan</th><th>Price</th><th>Pages / month</th><th>Rate limit</th><th>Highlights</th></tr>
<tr><td><b>Free</b></td><td>$0</td><td>30</td><td>1 req/min</td><td>All five operations, fast parse, no credit card</td></tr>
<tr><td><b>Starter</b></td><td>$1</td><td>300</td><td>10 req/min</td><td>Email support</td></tr>
<tr><td><b>Plus</b></td><td>$25</td><td>10,000</td><td>30 req/min</td><td>Advanced parse, team workspace with 3 seats</td></tr>
<tr><td><b>Pro</b></td><td>$99</td><td>50,000</td><td>60 req/min</td><td>Priority routing, up to 10 seats</td></tr>
<tr><td><b>Custom / Enterprise</b></td><td>Talk to us</td><td>Volume</td><td>Custom throughput</td><td>BYOK, IP allow-listing, SSO, human review step, custom pipelines, forward-deployed engineers</td></tr>
</table>
<p class="small" style="margin-top:12px">Current prices and promotions are on the <a href="https://apis.hyperbots.com/pricing" target="_blank" rel="noopener">pricing page</a>.</p>
</div>
</section>
<section class="alt" id="trust">
<div class="wrap two">
<div>
<h2>Security &amp; data handling</h2>
<ul class="clean">
<li>ISO 27001, SOC 1 Type 2 and SOC 2 Type 2, with encryption in transit and at rest</li>
<li>Documents are processed in memory and deleted when the run completes unless you enable retention</li>
<li>Customer data is never used for training. Redaction can run before any model sees a document</li>
<li>BYOK, IP allow-listing, SSO (Okta, Entra ID, Google Workspace, OneLogin), role-based access and full audit trails</li>
</ul>
</div>
<div>
<h2>The models</h2>
<p>HyperAPI runs on <b>HyperLM</b>, a mixture-of-experts document stack of small, vision and large language models, with multimodal encoders for layout, text and image and decoders specialised by task and field. It is trained on 35M+ real documents and 0.5B synthetic samples and retrained continuously.</p>
<p>Related research: <a href="https://www.hyperbots.com/research/breaking-the-annotation-barrier-with-doculite" target="_blank" rel="noopener">DocuLite</a> (4th Deployable AI Workshop, AAAI 2026) adapts small LLMs and VLMs to invoice extraction using synthetic data instead of human annotation.</p>
</div>
</div>
</section>
<footer>
<div class="wrap">
<div class="links">
<a href="https://apis.hyperbots.com" target="_blank" rel="noopener">apis.hyperbots.com</a>
<a href="https://apis.hyperbots.com/docs" target="_blank" rel="noopener">Docs</a>
<a href="https://github.com/hyprbots/hyperapi-sdk" target="_blank" rel="noopener">Python SDK</a>
<a href="https://apis.hyperbots.com/blogs" target="_blank" rel="noopener">Engineering blog</a>
<a href="https://apis.hyperbots.com/docs/status" target="_blank" rel="noopener">Status</a>
<a href="https://huggingface.co/hyperbots-inc" target="_blank" rel="noopener">Hyperbots on Hugging Face</a>
<a href="mailto:hyperapi-support@hyperbots.com">hyperapi-support@hyperbots.com</a>
<a href="mailto:research@hyperbots.com">research@hyperbots.com</a>
</div>
<p style="margin-top:16px">© 2026 Hyperbots Inc. HyperAPI is a product of Hyperbots Inc. Performance figures are as reported by Hyperbots; results on your documents may differ.</p>
</div>
</footer>
</body>
</html>