Scrape public websites and collect structured data through HasData MCP. Use the results in the files, reports and apps you build with OpenAI Codex.
1,000 free credits every month. No credit card required.
Find coworking spaces in Austin and save their details to data/places.json.
Worked for 45s
I’ll use HasData to find the spaces and save a JSON file.
HasData · Google Maps searchdata/places.jsonSaved 5 places with names, websites and ratings.
data/places.jsonJSON
// projects you can build
Start with a query, a location or a public page. Ask Codex to collect the data through HasData and work with it in your project.
Collect the pages ranking for a query. Save the results in your project and prepare a report you can share with your team.
Use HasData to get the first page of Google results for instagram profile api in Texas, United States. Save the returned positions, titles, URLs and snippets to data/serp.csv. Write reports/search-intent.md comparing the pages and their apparent search intent, with source links.
Start a directory with business names, websites and ratings from Maps. Ask Codex to add filters and a table to your app.
Use HasData to find the first page of coworking spaces in Austin, Texas. Save the returned names, addresses, websites and Google ratings to data/places.json. Build a directory view in this project with sorting by rating. Leave missing fields blank.
Collect listed prices, ratings and product links. Use the returned listings as the dataset for a comparison page.
Use HasData to find up to five USB-C docking stations on Amazon.com with one search. Save the returned product names, listed prices, currencies, ratings and URLs to data/products.json. Build a comparison table in this project. Mark unavailable fields as missing.
Read a public API reference and save it as Markdown. Use the retrieved documentation when Codex writes integration code in your project.
Use HasData to read this public API documentation page: [DOCUMENTATION URL]. Save the returned content to research/api-reference.md. Inspect this project's code and implement [INTEGRATION TASK] using the retrieved reference and existing project conventions. Cite the documentation URL and flag any missing endpoint or authentication details instead of guessing.
Add business addresses, phone numbers and service descriptions to an existing company CSV, with source links for the fields you collect.
Use HasData to enrich the first business record in data/companies.csv. Find its Google Maps listing using the supplied name and city, and read one public page from its supplied website. Check the name, location and website domain before merging. Write data/companies-enriched.csv with the original fields plus the returned address, business phone and a source-grounded description of services. Include available source URLs, leave missing values blank and flag uncertain matches.
Use Codex to turn retrieved sources into briefs, article drafts and social posts. Keep the workflow in your repo for repeatable runs.
Use HasData to get the first page of Google results for [TOPIC] and read up to two relevant public pages. Save the result URLs to research/content-sources.json and the retrieved page content as Markdown. Build a reusable workflow in this project that takes those source files as input and prepares a brief, an original article draft and three social post drafts under content/drafts/. Follow the project's existing tooling and style guidelines. Cite retrieved sources for factual claims and flag unsupported claims for review. Do not publish or schedule anything.
// without a custom scraper
Retrieve public page content as text or Markdown, or collect parsed records from supported sources. HasData handles proxies, source parsing and optional JavaScript rendering so you can use the data in your project.
HasDataYour Codex project
data/serp.csvdata/products.jsonreports/summary.mdWork with documented JSON fields and data types. Codex can map responses to your app's models and handle missing values explicitly.
Ask Codex to save JSON or CSV, write Markdown notes, or build a view from the retrieved records.
Use the same credit balance for Codex, other integrations and direct API requests.
// data sources
Google results, Shopping, News, Trends and cited answers from Google AI Mode.
Google Maps listings, place details and reviews. Yelp and YellowPages data too.
Amazon products and sellers, Walmart listings and public Shopify catalogs.
Page text and Markdown from public URLs, with JavaScript rendering when you need it.
YouTube videos and transcripts, public Instagram content and TikTok results.
Indeed, Glassdoor, Zillow, Redfin and supported flight and accommodation listings.
// setup
Follow these steps in the Codex IDE extension, or use the CLI commands below.
Open the gear menu → MCP servers → Add server. Name it HasData, choose Streamable HTTP and paste the URL below.
https://mcp.hasdata.com/mcpSave and select Restart extension. Then select Authenticate, sign in to HasData and authorize the connection.
Create a free HasData accountOpen your project and describe the data you need. Tell Codex what to collect and where to save it.
Choose a sample promptWith Codex CLI installed, run these commands and complete the HasData sign-in in your browser.
codex mcp add hasdata --url https://mcp.hasdata.com/mcp
codex mcp login hasdata// what it costs
Connecting HasData has no separate fee. Successful data requests use your shared HasData balance. Your Codex plan is separate.
One Google SERP Light request for a query.
5 credits per request
One Google Maps search, which can return multiple businesses.
10 credits per request
One Amazon search, which can return multiple product listings.
10 credits per request
Each example uses the full free balance for that request type. Extra tool calls use additional credits. See HasData plans
// FAQ
Need help with your connection or a data source? See the HasData docs or ask on Discord.
Yes. Connect HasData through MCP, give Codex a public URL and describe the data you need. Codex can retrieve page content through HasData and use it in a file, a report or your application. HasData runs the scraping on its hosted infrastructure.
Add https://mcp.hasdata.com/mcp as a remote MCP server and authenticate with your HasData account. In Codex CLI, use codex mcp add hasdata --url https://mcp.hasdata.com/mcp, then codex mcp login hasdata. The setup section includes copyable commands.
Codex web search helps with web lookups. HasData provides source-specific datasets such as Google result positions for a chosen location, Maps business details and marketplace product fields. Use those records to build a table, compare listings or prepare data for an app.
Yes. Ask Codex to write the retrieved records to CSV or JSON, or summarize page content in Markdown. Codex writes the files in a session with access to your project, subject to its permissions. HasData retrieves the source data and does not modify your repository.
No. Connect your Codex client to the hosted HasData MCP endpoint. You do not need to run a local scraping service or maintain parsing code for the supported data sources. Use a Codex client that supports remote MCP servers.
There is no separate HasData connection fee. Successful data requests use your HasData credit balance, and new accounts receive 1,000 free credits every month. Your Codex or ChatGPT plan is separate. A prompt that uses several HasData tools consumes credits for each successful request.
This integration is for public web data. It does not provide access to private social accounts or content behind a login. Use public page URLs and supported sources for the dataset you want to build.