Mojeek Focus Search Scraper

What is this tool?

Mojeek Focus Search Scraper lets you collect search data from Mojeek Focus — a feature of the privacy-respecting Mojeek search engine that lets you search within selected sets of websites. You can use built-in Focus templates (Recipes, TechNews, Education, etc.) or define your own custom site lists.

What can you use it for?

• Search within curated website collections (e.g., only news sites, only educational domains)
• Niche research across specific industry or topic sites
• Monitor how topics appear within selected authoritative sources
• Build focused datasets from trusted domains

What data do you get?

Each result includes source-backed search and focus fields. Optional fields are omitted when Mojeek does not expose them:

Identity (4 fields): Type, Result Type, Position, Page

URLs (8 fields): URL, Link (alias), Display URL, Domain, Host, Favicon, Image URL, Search URL

Text (5 fields): Title, Title (HTML), Snippet, Snippet (HTML), Description

Media (2 fields): Thumbnail, Images

Dates & Duration (4 fields): Date, Published At, Published At (raw), Relative Time

Categorization (2 fields): Tags, Category

Search Context (4 fields): Language, Search Metadata, Query, Query (alias)

Focus Context (3 fields): Focus Template, Focus Type, Focus Sites

Other (3 fields): Ad, Sponsored, Scraped At

How to use it

Step 1: Enter your search

Type in your search term, like "artificial intelligence".

Step 2: Choose your Focus

Pick a built-in template like "TechNews" or "Recipes", or provide a custom JSON array of domains.

Step 3: Set your preferences

Adjust the options below to customize your search.

Step 4: Run the tool

Click the run button and wait for your results.

Input options

Option	What it means	Required	Default
Search Query	Optional primary query. Runs first when combined with Queries.	No, if Queries is supplied	artificial intelligence
Queries	Optional batch of queries. The primary query is prepended and duplicates are removed.	No, if Search Query is supplied	—
Focus Template	Built-in Focus template name (Recipes, TechNews, Education, etc.).	No	Recipes
Focus Site Domains	Custom domain list, e.g. ["wikipedia.org", "britannica.com"]. Overrides Focus Template when non-empty.	No	—
Custom Focus Sites	Legacy JSON-array string accepted as an alias for Focus Site Domains.	No	—
Max Items	Maximum results across the full run.	No	50
Max Pages	Maximum result pages requested per query.	No	3
Max Concurrency	Browser pages processed at once.	No	2
Max Request Retries	Retries temporary network failures only (0–2). HTTP 403, HTTP 429, and detected challenge pages fail immediately without retrying.	No	1
Language	Language code for results (e.g. 'en', 'de', 'fr', 'es').	No	en
Safe Search	Safe search level: off, moderate, or strict.	No	0
Proxy Configuration	Optional Apify proxy settings. A proxy does not bypass Mojeek access restrictions.	No	Apify Proxy on platform

The complete input schema, including current and compatibility fields, is in .actor/input_schema.json.

Access limitations

https://www.mojeek.com/robots.txt (unchanged since 2024-02-21) publishes User-agent: * with Disallow: /search, Disallow: /url, Disallow: /focus/dashboard, Disallow: /eval and Disallow: /image?. The Actor's only query route is /search (Focus is applied with foc_t= / foc=), so it is disallowed for every automated client independently of the human-verification gate that sits on the same route.

As observed on 2026-10-25, that route refuses automated requests two ways: a headless Chromium client receives HTTP 403 with the title 403 - Forbidden ("Sorry your network appears to be sending automated queries…"), while a plain Chrome user agent receives HTTP 200 with the title Captcha ("JavaScript is required to complete this challenge…"). The challenge page injects <altcha-widget challenge="/captcha/challenge"> with no auto attribute, so it verifies only on a human checkbox click and never resolves on its own. There is no ungated Mojeek-owned Focus surface: https://www.mojeek.com/focus/ is a landing page whose only form posts to /search, /focus/dashboard is a settings page with no result cards (and is itself disallowed), and /focus/search and /search/ are 404.

The Actor reports what it observed and fails closed instead of returning an empty success:

• the run evaluates robots.txt once and logs the matched rule;
• the rendered page is read before the status code, so the HTTP 200 challenge page and the HTTP 403 page are both classified from their actual content;
• RATE_LIMITED (HTTP 429), TARGET_VERIFICATION (challenge page), TARGET_BLOCKED (HTTP 403) and TARGET_DISALLOWED (robots rule with no result cards) are terminal — no retry, no fingerprint rotation, no challenge solving;
• the key-value store OUTPUT record carries status, errorCode, httpStatus, pageUrl, pageTitle, pageExcerpt, resultCards, robotsMatchedRule and recordCount so a refusal is self-describing.

For compatibility, queries and focusSiteDomains are accepted alongside the local query and focusSites inputs. A non-empty focusSiteDomains list takes precedence; when it is empty, the legacy JSON-array focusSites value is used.

Available Focus Templates

These are the 15 templates the live https://www.mojeek.com/focus/ page offers:

Cybersecurity, Dictionary, Education, Encyclopedias, FactCheckers, GameDevAssets, Passwords, PrivacyAuthorities, PrivacyBlogs, Recipes, SmallWeb, TechNews, Texas, Time, Weather

News and the extra privacy sub-templates are not offered by the live landing page, so they are no longer advertised here: selecting one would silently produce an unfiltered /search request.

Output

Your results come in a clean, organized table with two views:

Overview view

Shows the most important fields at a glance:

• Position, Title, Domain, Snippet, Focus Template, Query, Scraped At

Full details view

Shows all 36 fields in detail.

The complete dataset field contract is in .actor/dataset_schema.json.

Tips for best results

✅ Use a relevant Focus template to narrow results to trusted sources

✅ Define custom focus sites for precise control over which domains are searched

✅ Be specific with your search terms for more relevant results

Frequently asked questions

What is Mojeek Focus? Mojeek Focus is a feature that lets you search within selected sets of websites. It works with Mojeek's own search index and requires no login or signup.

How do I use a custom focus? Provide a JSON array of domain strings in the "Custom Focus Sites" field, e.g. ["wikipedia.org", "britannica.com"]. This overrides the focusTemplate setting.

How do I get the results? Your results are saved in your Apify dashboard. You can download them in various formats (JSON, CSV, Excel) or access them via the API.