---
title: "How to use Retriever AI with your AI agent | ZeroClick"
url: https://zeroclick.ai/resources/how-to-use-retriever-ai-with-your-ai-agent
fetched_at: 2026-11-04T14:21:55.629Z
---

# 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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/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/qrpyfc3b7b4l/SKILL.md) or [full direct workflow reference](https://agents.zeroclick.ai/zcj/qrpyfc3b7b4l/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`.

# How to use Retriever AI with your AI agent

*Published October 28, 2026 · 9 min read · [ZeroClick](https://x.com/zeroclick)*

Retriever AI loads web pages in real browsers for your AI agent. For each URL, it returns the visible text, an accessibility tree with a number on every link, and a map from those numbers to the link targets. The price is a flat $0.05 per call. Retriever AI sells through ZeroClick: your agent sends its request to the [Retriever AI storefront](https://agents.rtrvr.ai), pays from a USDC wallet over x402 or MPP, and gets one tab back per URL, with no signup and no API key.

**Contents**
- [What Retriever AI is](#what-retriever-ai-is)
- [What agents use Retriever AI for](#what-agents-use-retriever-ai-for)
- [Prices](#prices)
- [How to buy and call Retriever AI from your agent](#how-to-buy-and-call-retriever-ai-from-your-agent)
- [Example requests](#example-requests)
- [If your agent has no wallet](#if-your-agent-has-no-wallet)
- [What to know](#what-to-know)
- [Frequently asked questions](#frequently-asked-questions)

**In short**
- Send `POST https://agents.rtrvr.ai/scrape` with a `urls` list, pay the 402 that comes back, and retry the identical request with the proof.
- A call costs $0.05 whether it carries one URL or several, so batching pages into one call keeps the cost down.
- Set `onlyTextContent` when a pipeline needs text alone, and add a webhook when results should be pushed to your own endpoint.
- An agent without a wallet can create one at no cost, or hand the purchase to a person through a tracked link.

---

## What Retriever AI is

Retriever AI builds an AI browser agent that works across the open web and sites you are signed into, then returns the finished result. It runs as a free Chrome extension, on Retriever AI's cloud of browsers, and inside other products through its API, MCP server, CLI, and SDK.

The service it sells to agents is the reading layer underneath: Scrape, which runs on Retriever AI's cloud browser cluster. Each page opens in a real browser behind rotating proxies, so pages that a plain HTTP fetch cannot read still come back. No language model runs on the way out, so your agent gets raw page data to search, extract from, or index with its own tools.

That suits an agent's work. An agent asked to read a set of pages and summarize them needs each page as a browser renders it, in a structure it can parse, within the same session it was given the task. The accessibility tree with numbered links gives it a compact picture of the page and a direct way to choose which link to follow next.

An agent storefront is the machine-readable version of a business. At this URL an agent reads what is for sale, the live price, and the payment steps, then buys without anyone filling in a form. It sits in front of Retriever AI's own API as a transparent proxy that handles payment and forwarding. For how storefronts work, see [What is an agent storefront?](https://zeroclick.ai/resources/what-is-an-agent-storefront)

---

## What agents use Retriever AI for

Every job below goes through the one Scrape service at `POST /scrape`. What changes is the body.

| Job your agent is given | What to send | What comes back |
|---|---|---|
| Read a web page as a browser shows it, for research | `urls` with one URL | One tab with `title`, `tree`, and `elementLinkRecord` |
| Read a list of pages in one go | `urls` with several URLs, optionally `maxParallelTabs` | One tab per URL, in the order sent |
| Feed text into a retrieval or RAG pipeline | `settings.extractionConfig.onlyTextContent: true` | Text only, without the tree and link map |
| Find the links on a page and pick the next one to follow | Default settings | Numbered link ids in `tree`, mapped to targets in `elementLinkRecord` |
| Read a page that loads more content as you scroll | Default settings, which include a scroll pass | Lazy-loaded content triggered before extraction |
| Send a scrape's result to your own endpoint | `webhooks` with an HTTPS `url` | A JSON envelope POSTed on success, failure, or cancellation |

`disableAutoScroll` skips the scroll pass when your agent does not need it. `pageLoadDelay` adds a wait after load before extraction. `trajectoryId` groups related scrapes for your own analytics. The [Scrape API reference](https://agents.rtrvr.ai/services/scrape/openapi.json) lists every field.

---

## Prices

| Service | Plan | Price | What a call covers |
|---|---|---|---|
| Scrape | Pay per request | $0.05 per call | One or more URLs, one tab per URL |

The flat rate makes batching the main lever on cost. Reading 50 pages one per call costs $2.50. Sending the same 50 pages as ten calls of five URLs costs $0.50. The plan includes no free units, and every call is charged as one request.

Prices come from the live [price list](https://agents.rtrvr.ai/manifest.json), and every 402 states the exact amount before your agent pays. Read the price at call time instead of caching it.

---

## How to buy and call Retriever AI from your agent

1. Point your agent at the [Retriever AI storefront](https://agents.rtrvr.ai), whose [agent guide](https://agents.rtrvr.ai/llms.txt) links the services, the price list, and the payment steps.
2. If the task is unclear, ask the free task guide at `https://agents.rtrvr.ai/zeroclick/agent/guide?goal=URL_ENCODED_TASK`, which needs no wallet and never buys anything.
3. Build the JSON body with the `urls` to read and any settings the job needs, such as text only or a webhook.
4. Send `POST /scrape` with no payment header. The storefront answers with a priced 402. Check `payment.amountUsd` against your budget, pay it from your wallet (x402 for USDC on Base, MPP for USDC on Tempo), and retry the identical request with the proof attached. The payment is bound to the body it priced.
5. Read the `tabs` array in the order you sent the URLs, and check each tab's `status` before using its content.

---

## Example requests

Send the unpaid request first. This one asks for two pages in a single call:

```http
POST https://agents.rtrvr.ai/scrape
Content-Type: application/json

{"urls": ["https://example.com", "https://www.iana.org/help/example-domains"]}
```

The 402 body carries `payment.amountUsd`, the `plan`, and a `protocols` block with a challenge for each rail. The signable challenges are repeated in the response headers, which lets an off-the-shelf x402 or MPP client pay without parsing the body:

```http
HTTP/1.1 402 Payment Required
payment-required: <x402 challenge>
www-authenticate: <MPP challenge>
```

Sign the challenge and resend the same request with the proof. On x402 the payload goes in `x-payment`. On MPP, send `Authorization: Payment <credential>` instead:

```http
POST https://agents.rtrvr.ai/scrape
Content-Type: application/json
x-payment: <signed x402 payload>

{"urls": ["https://example.com", "https://www.iana.org/help/example-domains"]}
```

The paid response carries `success`, a `tabs` array with one entry per URL, `metadata`, and `usageData`, with a `zc-billing` header that reconciles `authorizedUsd` against `chargedUsd`. Here is part of the response from a single-page call, paid through Zero (ZeroClick's own payment layer for agents), trimmed to the fields an agent reads first:

```json
{
  "success": true,
  "tabs": [
    {
      "url": "https://example.com/",
      "title": "Example Domain",
      "contentType": "text/html",
      "status": "success",
      "tree": "[textNode] This domain is for use in documentation examples without needing perm…",
      "elementLinkRecord": {"8": "https://iana.org/[path]"}
    }
  ]
}
```

A single `https://example.com` call cost $0.05, took about 10 seconds, and returned one tab. The tree contained the page's text as text nodes, and the link map pointed element 8 at an iana.org address. A two-URL call also cost $0.05, took about 16 seconds, and returned two tabs.

---

## If your agent has no wallet

- **Create a wallet.** Your agent can set one up itself, at no cost and with no signup, then follow the steps above.
- **A card does not fit here.** A card pays only a call of $0.50 or more. A Scrape call costs $0.05, and the catalog lists no credit plan to fund by card, so calls are paid from a wallet.
- **Hand off to a person.** Share any Retriever AI page as `https://agents.rtrvr.ai/outclick?url={URL-encoded seller page}` so the visit is recorded as coming from an agent, and let whoever approves the purchase take it from there.

---

## What to know

**What the service does not do.** Scrape returns page data and nothing more: no summaries, no extracted fields, and no model output. The storefront sells this one service.

**What is charged, and when.** Each call is one 402 and one charge of $0.05, however many URLs it carries. Paying the 402 completes the purchase under ZeroClick's [buyer terms of service](https://www.zeroclick.ai/legal/buyer-terms-of-service). `usageData` in the response describes Retriever AI's infrastructure use and is informational only. A delivered response with no `zc-billing` header billed nothing.

**Large pages.** Inline output is capped at 1,048,576 bytes by default, set with `response.inlineOutputMaxBytes`. A larger result returns per-tab previews, with `metadata.responseRef` pointing at the full payload.

**Webhooks.** Each webhook takes an HTTPS `url`, optional bearer or basic `auth`, a `secret` that adds an `X-Rtrvr-Signature` HMAC header, and a `timeoutMs` that defaults to 8,000. By default it delivers `rtrvr.scrape.succeeded`, `rtrvr.scrape.failed`, and `rtrvr.scrape.cancelled`, and it retries with backoff unless `retry.mode` is `none`.

**Errors worth handling.** A tab can fail while the call succeeds, so check each tab's `status` and read its `error` when the status is `error`. If the retry's body differs from the priced request, the storefront refuses it with `409 payment_request_mismatch`; request a fresh 402 and pay that. The [error reference](https://agents.rtrvr.ai/errors.md) lists every code.

---

## Frequently asked questions

### Does Retriever AI need an API key for AI agents?

No. Through the ZeroClick storefront your agent pays each call with a signed payment, which also identifies it, so no signup or API key is needed.

### How much does Retriever AI cost per page?

A Scrape call costs a flat $0.05 and can carry one or more URLs, so five pages in one call cost $0.05 in total.

### What does the Retriever AI API return?

For each URL it returns one tab with the text a reader sees on the page, an accessibility tree that numbers every link, a map from those numbers to link targets, and the page title, in the order the URLs were sent.

### Can Retriever AI return text only, without the accessibility tree?

Yes. Set `settings.extractionConfig.onlyTextContent` to `true`, and each tab returns the extracted text without the tree or the link map.

### Do I have to use an AI agent to use Retriever AI?

No. A person can use Retriever AI directly from [rtrvr.ai](https://rtrvr.ai), where its free, ad-supported browser agent runs as a Chrome extension, with Cloud, an API, MCP, a CLI, and an SDK for running it at scale or inside a product.

---

## More in: Get more done with your agent

- [How to use Keenable with your AI agent](https://zeroclick.ai/resources/how-to-use-keenable-with-your-ai-agent) — Keenable sells web search, page fetches, and SQL over the web to AI agents through ZeroClick, at $0.004 a call and $0.10 a SELECT, on Credits from $0.50.
- [How to use Pubrio with your AI agent](https://zeroclick.ai/resources/how-to-use-pubrio-with-your-ai-agent) — Pubrio sells market expansion signals and company data to AI agents through ZeroClick, from $0.01 per credit, paid per call with no signup or API key.
- [How to use Zenrows with your AI agent](https://zeroclick.ai/resources/how-to-use-zenrows-with-your-ai-agent) — Zenrows sells web page fetches to AI agents through ZeroClick, from $0.50 per 1,000 credits, one credit per plain page, and failed requests cost nothing.

[All resources](https://zeroclick.ai/resources#get-more-done)
