# AI Web Research with Cited Sources (`patchwell/patchwell-research`) Actor Ask questions in plain English and get answers where each sentence cites the page it came from, plus what the sources couldn't confirm and a confidence grade. Up to 50 questions per run. You pay $0.15 only for answered questions. - **URL**: https://apify.com/patchwell/patchwell-research.md - **Developed by:** [Aidan McCollum](https://apify.com/patchwell) (community) - **Security:** Limited permissions - **Stats:** 2 total users, 1 monthly users, 100.0% runs succeeded, 0 bookmarks - **User rating**: No ratings yet ## Pricing from $150.00 / 1,000 answered questions This Actor is paid per event. You are not charged for the Apify platform usage, but only a fixed price for specific events. Learn more: https://docs.apify.com/actors/running/actors-in-store.md#pay-per-event ## What's an Apify Actor? An Actor is a serverless cloud program that runs on the Apify platform. It has two run modes. In Batch mode, an Actor accepts a well-defined JSON input, performs an action which can take anything from a few seconds to a few hours, and optionally produces a well-defined JSON output, datasets with results, or files in key-value store. In Standby mode, an Actor provides a web server which can be used as a website, API, or an MCP server. Apify vocabulary and the platform model are defined once, in the agent quickstart at https://apify.com/agents.md. ## How to integrate an Actor? If asked about integration, you help developers integrate Actors into their projects. You adapt to their stack and deliver integrations that are safe, well-documented, and production-ready. Do not guess an integration path. Every one of them is in the agent quickstart at https://apify.com/agents.md: the Apify MCP server, Agent Skills with the Apify CLI, the JavaScript and Python clients, the REST API, and the account-free path for an agent with no human to sign in. It also carries the rule on stating cost before the first paid run. For examples already wired to this Actor's own input schema, see the [API](#api) section below. Each client library has reference documentation the quickstart does not restate: [JavaScript/TypeScript](https://docs.apify.com/api/client/js/docs.md) (`npm install apify-client`) and [Python](https://docs.apify.com/api/client/python/docs.md) (`pip install apify-client`). # README Ask a question in plain English and get an answer where **each sentence cites the page it came from**. The answer also lists what the sources couldn't confirm and gives a confidence grade, so you can judge how far to trust it. Run one question or a batch of 50. **You pay only for questions that get an answer.** ### What you get for each question | Field | What it holds | |---|---| | `answer` | The answer. Each sentence ends with the URL of its source in brackets | | `sources` | The pages the answer relies on, with their titles | | `gaps` | What the sources couldn't confirm, and how many claims the checker dropped | | `confidence` | `high`, `medium` or `low`, based on how many claims held up against their sources | | `answered` | `false` when no source backed an answer. You aren't charged for these | | `searches` | The web searches it ran | #### How it checks its own answer The research engine searches the web, reads the pages and drafts an answer. A separate step then compares each claim with the text of the page it cites, drops the claims that page doesn't support, and counts them in `gaps`. You can check any remaining sentence in one click. Sources can disagree with each other, so for anything important, open the cited page. ### How to use it 1. Type your questions into **Questions**, one per line. 2. Add **Context** if it applies to all of them, such as who the answer is for. 3. Click **Start**. The **Answers** tab in Output shows one row per question. **Sources and gaps** shows what each answer rests on. A question takes about a minute, and two run at a time. After a quiet period, the first question can take up to a minute longer while a GPU starts. You can also call it through the [Apify API](https://docs.apify.com/api/v2), run it on a schedule, or connect it to Make, Zapier and n8n. ### What to ask it Questions about **markets, products, prices, companies, regulations, places, events and published research**. For example: - What do the top three payroll providers charge a 20-person company? - Which EU countries require e-invoicing for B2B sales in 2026? - What did recent studies find about four-day work weeks and productivity? It **won't look up private people**: no addresses, phone numbers, relatives or profiles of individuals. It declines those questions and doesn't charge for them. ### How much does it cost? **$0.15 per answered question.** Declined questions and questions without a sourced answer cost nothing. | Questions answered | Cost | |---|---| | 1 | $0.15 | | 50 (one full batch) | $7.50 | | 1,000 | $150.00 | ### Example Input, using the question prefilled in the form: ```json { "questions": ["What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?"] } ``` Real output from that run, which took about a minute (3 of its 8 sources shown): ```json { "question": "What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?", "answered": true, "confidence": "medium", "answer": "Stripe charges 2.9% + 30 cents for standard online domestic card payments in the US [https://www.swipesum.com/insights/guide-to-stripe-fees-rates-for-2025][https://eco.com/support/en/articles/17338153-stripe-vs-square-2026-fees-payouts-and-which-fits-your-business][https://checkoutpage.com/blog/stripe-processing-fees]. Stripe and Square charge the same 2.9% + 30ยข for standard online card payments [https://orbitforms.ai/blog/stripe-vs-square-vs-paypal]. Square charges 2.9% + $0.30 for online payments through Square APIs [https://www.swipesum.com/insights/square-fees-explained-understanding-your-payment-costs]. PayPal's standard merchant fee is 3.49% + 49ยข per online transaction [https://merchantinsiders.com/blogs/paypal-fees/][https://orbitforms.ai/blog/stripe-vs-square-vs-paypal]. PayPal Checkout fees for online transactions are 2.99% plus 49 cents for debit and credit card payments [https://www.nerdwallet.com/business/software/learn/paypal-fees][https://merchantinsiders.com/blogs/paypal-fees/].", "sources": [ { "url": "https://www.swipesum.com/insights/guide-to-stripe-fees-rates-for-2025", "title": "Stripe Fees & Rates 2026" }, { "url": "https://eco.com/support/en/articles/17338153-stripe-vs-square-2026-fees-payouts-and-which-fits-your-business", "title": "Stripe vs Square 2026: Fees, Payouts and Which Fits Your Business - Eco" }, { "url": "https://checkoutpage.com/blog/stripe-processing-fees", "title": "Stripe fees 2026" } ], "gaps": [ "2 claims left out because the sources found did not clearly support them." ] } ``` ### Use it from an AI agent Add this Actor to Claude, Cursor or any MCP client through [Apify's MCP server](https://mcp.apify.com). With a citation on each sentence, you can check any claim your agent repeats. ### FAQ #### Which model does it use? Qwen3.6, an open-weight model, on GPUs Patchwell rents, plus the separate checking step described above. Your question stays off outside AI companies' servers. The short search queries the model writes do go to web search engines. #### Is my question stored? By default, Patchwell keeps the question text and basic quality signals for 30 days, with contact details removed, and uses them to improve accuracy. Turn off **Let Patchwell keep questions** in the input and Patchwell keeps nothing. Results stay in your own Apify storage. #### Why was my question declined? It asked about a private person, or tried to get the tool to ignore its rules. Ask about the topic instead of the person. You don't pay for declined questions. #### Can it read a specific page or PDF I give it? Not yet. It finds its own sources on the open web. ### Other Patchwell tools - [Audio & Video Transcriber](https://apify.com/patchwell/patchwell-transcribe): word-level timestamps, SRT and VTT for audio and video links. - [CSV & Excel Data Cleaner with Change Log](https://apify.com/patchwell/patchwell-clean-data): clean messy spreadsheets or any Actor's dataset, with each fix logged. Questions or a custom need: patchwell.support@gmail.com # Actor input Schema ## `questions` (type: `array`): One question per line, up to 50 per run. Ask about topics, markets, products, companies, places, events and published research. Two run at a time. It won't look up private people. ## `context` (type: `string`): Background that applies to every question, such as who the answer is for or what you already know. Up to 4,000 characters. ## `keepQuestions` (type: `boolean`): When on, Patchwell keeps the question text and basic quality signals for 30 days, with contact details removed. Turn it off and Patchwell keeps nothing. ## Actor input object example ```json { "questions": [ "What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?" ], "keepQuestions": true } ``` # Actor output Schema ## `answers` (type: `string`): No description ## `sources` (type: `string`): No description # API You can run this Actor programmatically using our API. Below are code examples in JavaScript, Python, and CLI, as well as the OpenAPI specification and MCP server setup. ## JavaScript example ```javascript import { ApifyClient } from 'apify-client'; // Initialize the ApifyClient with your Apify API token // Replace the '' with your token const client = new ApifyClient({ token: '', }); // Prepare Actor input const input = { "questions": [ "What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?" ] }; // Run the Actor and wait for it to finish const run = await client.actor("patchwell/patchwell-research").call(input); // Fetch and print Actor results from the run's dataset (if any) console.log('Results from dataset'); console.log(`๐Ÿ’พ Check your data here: https://console.apify.com/storage/datasets/${run.defaultDatasetId}`); const { items } = await client.dataset(run.defaultDatasetId).listItems(); items.forEach((item) => { console.dir(item); }); // ๐Ÿ“š Want to learn more ๐Ÿ“–? Go to โ†’ https://docs.apify.com/api/client/js/docs ``` ## Python example ```python from apify_client import ApifyClient # Initialize the ApifyClient with your Apify API token # Replace '' with your token. client = ApifyClient("") # Prepare the Actor input run_input = { "questions": ["What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?"] } # Run the Actor and wait for it to finish run = client.actor("patchwell/patchwell-research").call(run_input=run_input) # Fetch and print Actor results from the run's dataset (if there are any) print(f"๐Ÿ’พ Check your data here: https://console.apify.com/storage/datasets/{run.default_dataset_id}") for item in client.dataset(run.default_dataset_id).iterate_items(): print(item) # ๐Ÿ“š Want to learn more ๐Ÿ“–? Go to โ†’ https://docs.apify.com/api/client/python/docs/quick-start ``` ## CLI example ```bash echo '{ "questions": [ "What does Stripe charge for online card payments in the US, and how does that compare to Square and PayPal?" ] }' | apify call patchwell/patchwell-research --silent --output-dataset ``` ## MCP server setup ```json { "mcpServers": { "apify": { "type": "http", "url": "https://mcp.apify.com/?tools=fetch-actor-details,patchwell/patchwell-research" } } } ``` The hosted server signs you in with OAuth on first connect, so no API token belongs in this config. Clients without OAuth support can send an `Authorization: Bearer ` header instead, using a token from API & Integrations in Apify Console (https://console.apify.com/settings/integrations). ## OpenAPI specification Download the OpenAPI definition: https://api.apify.com/v2/actors/SDiqCryelqD6PhTjU/builds/wESpO6xoCcot9Lrvw/openapi.json