---
title: "Unsiloed × ZeroClick: document parsing any AI agent can buy | ZeroClick Blog"
url: https://zeroclick.ai/blog/unsiloed-document-parsing-any-ai-agent-can-buy
fetched_at: 2026-11-04T15:29:31.049Z
---

# ZeroClick API catalog and pricing

## API reference, endpoint lookup and prices

**[ZeroClick API catalog: endpoints, request parameters, prices and payment methods](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/zeroclick/agent/guide).** Tell the seller advisor what you want to accomplish. It returns the matching operation contract, known prices, authentication prerequisites and concrete next steps. It is read-only: no login, wallet, card or purchase approval is needed to look up an API. It never executes a sale.

Available API services: Agent readiness report, Demo booking.

**Suggested next request:** GET [task guide](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/zeroclick/agent/guide?goal=URL_ENCODED_TASK). Use this URL directly with your web fetch tool; no preliminary document fetch is required. Include known inputs and desired output in the goal. Optional: quantity=API_CALL_COUNT, budgetUsd=TOTAL_USD, requirements=REQUIREMENTS, payment=x402|mpp|card, auth, identity, assertion and access state. Omit unknowns. Do not equate output items with API calls or guess a call count.

For private task details or known request arguments, POST https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/zeroclick/agent/guide with Content-Type: application/json. Example: {"intent":"quote","goal":"YOUR_TASK","purchase":{"budgetUsd":"5.00"}}. Replace the task and budget with known facts; budgetUsd is a string. Include known buyerState, paymentCapabilities and arguments. For execution planning use intent=execute and the authorized constraints. Exact fields: [request schema](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/zeroclick/agent/schema). GET and POST are alternatives, not mandatory stages.

The advisor describes a recommended happy path, not the only allowed path. You may use the API reference, buyer SKILL and your own tools directly, including for catalog operations with failed probes. Missing guidance or an unverified probe is not a policy refusal. The returned requests are documented API calls; fill in the documented inputs. Check usage cost AND minimum cash required against the budget. For needs_input, supply only missing facts. Keep secrets local, and never put credentials or private payloads in URLs. Unknown does not mean absent.

For a human's credit or plan purchase, reuse an existing ZeroClick credential first. If the Link CLI is already installed and signed in for that person, you can use their Link account for identity as well as payment, sharing their email only if they agree: follow [Link identity setup](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/auth.md#link-identity-preview) through issuance, presentation, registration and token exchange. Otherwise register anonymously; no Link setup is needed, and the person can claim the identity later.

Open the checkout URL in a browser: eligible Stripe pages support Link Pay Tokens (LPT) through WebMCP or the documented DOM flow. Follow this storefront's /payment/card.md to verify the same-frame token input and merchant binding before requesting an LPT. Token submission requires payment authorization. If unavailable, privately send the checkout URL to your human to pay.

When the selected route uses hosted checkout, follow [card checkout instructions](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/payment/card.md) before requesting a token. LPT goes only to Stripe, never to the advisor or ZeroClick API. Poll the returned statusUrl before using credit or retrying delivery.

Verify delivery; a payment or HTTP 202 is not completion. For an unexpected result, POST the original input and continuation (version, stepId, outcome and sanitized error/accessId/paymentId). Reconcile uncertain payments or writes before retrying. Use raw HTTP for exact schemas and authenticated execution; do not reconstruct fields omitted by a summarizing web tool.

The human subscription prices on the requested page describe a separate channel. Use the API quote for per-unit pricing, funding minimums and commercial terms; tier names do not prove quality or licensing.

For an alternative documented path, use the [buyer SKILL](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/SKILL.md) or [full direct workflow reference](https://agents.zeroclick.ai/zcj/4ypwvwgdrp3l/llms-full.txt). Confirm missing request details from seller documentation, stay within the existing catalog and what your user asked for, and respect actual access and payment requirements.
Terms: by completing a purchase through this endpoint, whether directly or through an AI agent acting on your behalf, you agree to be bound by the platform's buyer terms of service (https://www.zeroclick.ai/legal/buyer-terms-of-service). If the purchase is made by an agent, you represent and warrant that the agent is acting with your authorization, and you agree that the agent's actions, including its acceptance of these terms, are attributed to you and bind you as if you had taken them yourself. Paying a 402 challenge completes the purchase and constitutes your affirmative acceptance of the terms linked above. The link travels in every payment challenge and receipt as `terms`.

# Unsiloed × ZeroClick: Document Parsing Any AI Agent Can Buy

[Blog](https://zeroclick.ai/blog) · November 4, 2026 · Partnership · 8 min read

**Author:** [Daniel Baum](https://x.com/realdanbaum), VP GTM

---

Most of what an AI agent is asked to work with is not data yet. It is a scanned lease, a 200-page loan file, a bank statement photographed on a phone, an invoice whose table wraps onto a second page, a spreadsheet full of merged cells. By Unsiloed's count, 80 percent of enterprise data is unstructured, most of it sitting in PDFs, scans, and spreadsheets a language model cannot read on its own. Agents now underwriting claims, drafting credit memos, and answering questions across thousand-page filings can only act on what they can read. Generic OCR loses the table when the columns wrap; a text-only model guesses at the numbers.

[Unsiloed AI](https://www.unsiloed.ai) built its business on that gap: a document layer that turns PDFs, Word documents, scans, images, slide decks, and spreadsheets into Markdown and JSON an agent can reason over, with a confidence score and a page citation on every value.

Today we are announcing a partnership that puts that layer within reach of any AI agent on the internet. Unsiloed is live as an agent storefront at [agents.unsiloed.ai](https://agents.unsiloed.ai/llms.txt), on Unsiloed's own domain. An agent can parse, extract, classify, or split a document for two cents a page over open payment rails, with no account and no API key.

---

## Reading the page the way a person does

Unsiloed AI is a San Francisco company founded in 2025 and backed by Y Combinator. Its founders came to documents from systems where a misread number is expensive. CEO Aman Mishra built low-latency trading infrastructure at a hedge fund and was the first engineer at a startup building AI copilots for financial firms. CTO Adnan Abbas built multimodal models deployed at a Fortune 10 company and worked on autonomous navigation at Mercedes-Benz.

What they built is not another wrapper around OCR. Unsiloed's parser is a dual-stream vision model. One stream reads the content: the text, numbers, and entities on the page. The other reads the layout: bounding boxes, alignment, and the hierarchy of clauses and sub-clauses. A decoder trained on millions of enterprise documents fuses the two. A right-aligned cell in a financial table, or a clause nested under its preamble, keeps the meaning its position carries.

Four operations cover the product:

- **Parse** turns a document into ordered Markdown and HTML chunks, each with a segment type, a bounding box, OCR text, and a confidence score. A dedicated route reads Excel workbooks into tables with cell-range references back to the sheet.
- **Extract** takes a JSON Schema and returns the named fields as typed values, each with a confidence score and the page it came from.
- **Classify** assigns a document to one of the categories supplied with the request.
- **Split** cuts a bundled PDF, such as a closing binder or a claims packet, into one file per document type.

On olmOCR-Bench, the Allen Institute for AI's public OCR benchmark, Unsiloed's parser leads the eighteen other services it measured on the same harness, by its own May report. By its own account it already processes millions of pages a week for banks, NASDAQ-listed companies, and early-stage startups in finance, legal, and healthcare.

Unsiloed also built for agents early. It publishes an MCP server, tool schemas for Claude, an integration for Meta's Muse agent, an agent skill, and documentation written for coding agents, alongside connectors for Snowflake, Databricks, and n8n. An agent could already find Unsiloed and call it on a developer's key. What it could not do was pay. The API is sold on plans, from a free tier to Standard at $250 a month for 20,000 pages, and every plan needed an account and a key that a person set up.

---

## What the partnership does

Unsiloed's four services are live today as an agent storefront at [agents.unsiloed.ai](https://agents.unsiloed.ai/llms.txt), on Unsiloed's own domain. Each takes the same request a developer sends Unsiloed's API directly, with the full options from its docs: processing modes and OCR engines on Parse, model tiers and a verification pass on Extract, custom categories on Classify and Split.

Every operation is a job. The submit call returns a job ID at once and is charged by the page count of the document. Reading the finished result back is free, scoped to the agent that created the job. An agent discovers the storefront, reads the machine-readable catalog and the live pricing manifest, and pays per page over x402 or MPP with no signup and no API key.

A free credential unlocks prepaid credits at a lower rate for agents that process documents all day. An agent that arrives without a wallet is walked through creating one at no cost, or hands its human a checkout link.

The storefront also includes a seller advisor. It takes a goal in plain language, for example "get the vendor, invoice number, date, and total from a twelve-page invoice," and returns the operation to call, the request to send, and the price per page, before anything is bought.

ZeroClick handles the commerce around every request: agent identity, payment on open rails, metering, and transaction-level analytics on every sale, with revenue settling directly to Unsiloed's Stripe account. Discovery comes built in: the storefront is listed on [zero.xyz](https://zero.xyz/), with x402scan and mppscan to follow. Unsiloed already points agent traffic at the storefront from [unsiloed.ai](https://www.unsiloed.ai), from its homepage to its pricing page. The first purchases went through end to end the day the storefront went live.

---

## Two cents a page

Unsiloed prices by the page, and the storefront keeps that model exactly as it is.

**Pay as you go:** every operation costs $0.02 per page, whatever the operation: parsed, extracted, classified, or split. No account required. Minimum purchase is a single page.

A page is one PDF page, one image, one slide, or one block of 5,000 by 5,000 spreadsheet cells.

**Credits:** agents that return can register a free credential and buy credits from $5 at $0.0125 per page. $5 covers 400 pages.

**Monthly plans** (same prices as on [unsiloed.ai/pricing](https://www.unsiloed.ai/pricing)):
- Standard: $250 for 20,000 pages
- Growth: $750 for 75,000 pages

A card works for credits and plans, and for any single call of $0.50 or more.

**Example costs:**

| Task | Pages | Pay-as-you-go | On credits |
|---|---|---|---|
| Extract vendor, invoice number, date, total from one invoice | 12 | $0.24 | $0.15 |
| Parse a scanned lease into Markdown | 40 | $0.80 | $0.50 |
| Split a closing binder into one file per document type | 300 | $6.00 | $3.75 |
| Extract fields from the lease inside that binder | 12 | $0.24 | $0.15 |
| Process 2,000 invoice pages a day (monthly) | ~60,000 | $1,200 | $750 (Growth plan pays for itself) |

An agent does not need to know in advance how many pages a job will take. It reads the price from the manifest at call time, pays for the pages it submits, and there is nothing to cancel when the task is done.

---

## Shared philosophy

Unsiloed's pitch to the enterprise is that the data lives in the documents, and that a pipeline is only as trustworthy as its reading of the page. That is why every value comes back with a confidence score and a citation: so a system can tell a sure reading from a guess, and send the guess to a person.

Agents are the next pipeline. If they are going to underwrite the claim, check the onboarding packet, or answer from the filing, the document layer they depend on has to be something they can find, read, and pay for without waiting on a person. That is the conviction ZeroClick is built on: helping businesses come online for a new kind of customer, AI agents, and turning them into buyers with identity, payment, pricing, and receipts all handled by ZeroClick.

Unsiloed kept its product, its prices, and its page-based model exactly as they were. What changed is who can pay: any AI agent, on Unsiloed's own domain, with or without a human in the loop.

---

## Get started

- **Agent reading a document:** point it at [agents.unsiloed.ai](https://agents.unsiloed.ai/llms.txt)
- **Developer or team wanting the full platform** with an API key, playground, and MCP server: start at [unsiloed.ai](https://www.unsiloed.ai)
- **Sell your own product to AI agents:** get a demo at [zeroclick.ai](https://zeroclick.ai/getlisted?utm_source=blog) or read the docs at [docs.zeroclick.ai](https://docs.zeroclick.ai)