MrScraper
This provider is an extra of the lfx-bundles metapackage. It is opt-in for both Langflow and standalone LFX installs:
uv pip install "lfx-bundles[<bundle>]"
Replace <bundle> with this page's provider name, for example qdrant.
For a torch-free full Langflow install, run uv pip install "langflow[bundles]".
For standalone LFX with every provider in lfx-bundles, including PyTorch-based providers, run uv pip install "lfx[bundles]".
See Additional bundles for the exact extra name.
Bundles contain custom components that support third-party integrations with Langflow.
The MrScraper bundle provides three components: Account, Scraper API, and Web Search. They use mrscraper-sdk version 0.6.x. Install the provider dependencies with pip install "lfx-bundles[mrscraper]", then enter your API token from the MrScraper app in each component's MrScraper API Token field.
All three components return the API payload as Data. The SDK's status_code and headers envelope is removed. Object responses are preserved; text and list responses are wrapped in a field so they remain valid Langflow data. API, authentication, and network errors fail the component.
MrScraper Account
Check account identity, subscription status, token quota, and rate limits. This component calls GET https://api.app.mrscraper.com/api/v1/subscription-accounts.
| Parameter | Default | Description |
|---|---|---|
MrScraper API Token (api_token) | Required | Your account's API token, stored as a secret. |
Timeout (seconds) (timeout) | 30 | Positive HTTP request timeout. |
The output includes subscription_status, enterprise, token_usage, token_limit, token_remaining, usage_percent, rate_limit, rate_ttl, auto_renew, ends_at, and user with name, email, and verified. Remaining tokens are clamped to zero when usage exceeds the limit. API keys and billing identifiers are excluded.
For example, connect Account to a Parser to read token_remaining before starting a scraping workflow.
MrScraper Scraper API
Fetch HTML, Markdown, screenshots, or a combination. This component sends POST https://api.mrscraper.com. The request body contains url, maxRetries, and optional tokenCap; page options and the API token are query parameters.
| Parameter | Default | Description |
|---|---|---|
MrScraper API Token (api_token) | Required | Your API token. |
URL (url) | Required | Target page URL. Available in Tool Mode. |
HTML (html) | true | Request HTML. |
Markdown (markdown) | false | Request Markdown, independently of HTML. |
Screenshot (screenshot) | false | Capture a screenshot; always enables browser rendering. |
Browser Rendering (browser_rendering) | Auto | Auto tries without a browser before falling back. Enabled forces rendering; Disabled disables it unless Screenshot is enabled. |
Super (super) | false | Enable real device routing independently of browser rendering. |
Timeout (seconds) (timeout) | 120 | Page timeout from 1 to 300 seconds. The SDK adds 30 seconds for the HTTP deadline. |
Geo Code (geo_code) | US | Geolocation country code. |
Proxy Country (proxy_country) | Empty | Optional proxyCountry value. |
Use Proxy (use_proxy) | true | Allow the API to use a proxy. |
Custom Proxy (proxy) | Empty | Optional secret proxy in username:password@host:port form. |
Block Resources (block_resources) | false | Block images, CSS, and fonts. |
Bypass Proxy (bypass_proxy) | false | Enable the API's proxy bypass behavior. |
Wait for Selector (wait_for_selector) | Empty | Wait for a CSS selector. |
Visit Home Page First (home_page) | false | Visit the site's root first; sent as warmStart=true. |
Cookie Jar (cookie_jar) | Empty | Optional configuration such as domain=example.com&country=us&max_used=100&max_age=50m. |
Target Headers (headers) | Empty | Target request headers as a JSON object string. |
Max Retries (max_retries) | 3 | Nonnegative maximum retries; 0 is preserved. |
Token Cap (token_cap) | Empty | Optional nonnegative integer retry budget; blank leaves it unset and 0 is preserved. |
Stream (stream) | Auto | Auto leaves streaming unset; Enabled or Disabled sends an explicit value. The SDK buffers the response. |
Discard Token Secret (discard_token_secret) | Empty | Optional secret x-discard-token-secret request header. |
Only the URL and output options are shown initially; open the component's advanced settings for the remaining controls. To retrieve all representations, enter https://example.com and enable HTML, Markdown, and Screenshot. Screenshot also enables browser rendering when Browser Rendering is Disabled.
JSON object responses pass through unchanged, including combined HTML, Markdown, and screenshot responses. Text responses are returned as {"html": "..."}; the html field name is retained for existing flows and can contain Markdown when only Markdown is requested. Other non-object payloads also use the html field. Inspect the response before assuming it contains an HTML string.
MrScraper Web Search
Search Google using query text or a full Google search URL. This component sends POST https://sync.scraper.mrscraper.com/api/google/serp/v2/sync.
| Parameter | Default | Description |
|---|---|---|
MrScraper API Token (api_token) | Required | Your API token. |
Query or Google URL (url) | Required | Query text or Google search URL. Available in Tool Mode. |
Raw HTML (raw) | true | Return HTML regardless of Format. Disable this to request structured JSON. |
Format (format) | json | json or html, used when Raw HTML is disabled. |
Region (region) | Empty | Optional country code; overrides gl from a URL. |
Language (language) | Empty | Optional language code; overrides hl from a URL. |
Page (page) | Empty | Optional positive integer page; overrides start from a URL. |
Render JavaScript (render_js) | false | Render JavaScript-dependent search features. |
Timeout (seconds) (timeout) | 600 | Positive HTTP request timeout. |
The SDK extracts q, gl, hl, and start from Google URLs. For example, https://www.google.com/search?q=running+shoes&gl=us&hl=en&start=20 searches for running shoes in region us, language en, on page 3. Explicit component settings override URL values.
For structured results, enter python web scraping, disable Raw HTML, and select Format: json. JSON objects pass through unchanged; HTML, list, and other scalar responses are wrapped as {"result": ...}.
Updating existing flows
MrScraper Fetch Rendered HTML is now displayed as MrScraper Scraper API. Its class and saved-flow identifier remain MrscraperFetchHtml, its output name remains data, and plain text output keeps the html field. Existing URL, timeout, geolocation, and resource-blocking settings remain available. Browser rendering now defaults to the SDK's automatic fallback; select Enabled to always use a browser.
The previous AI Agent Scraper, Batch Scrape URLs, Crawl Website, Get Result Detail, Get Results, Run AI Scraper, and Run Manual Scraper components have been removed. Their historical migration entries remain for old flow identification, but those operations have no equivalent in this three-component bundle. Replace affected nodes with Scraper API for page retrieval or Web Search for discovery, and update downstream parsing. A prompt-driven extraction workflow can connect page content to a language model; it requires rebuilding the workflow rather than an automatic node conversion.
For more information, see the MrScraper documentation.
Was this page helpful?
