# Siftwright

> Web data APIs for developers and AI agents: YouTube transcripts, website screenshots and PDFs, and Google News results, behind one API key. Only successful results are billed.

- Base URL: https://siftwright.com/v1 (POST JSON, Bearer auth with a key like sw_live_...)
- Endpoints: /v1/youtube-transcript, /v1/screenshot, /v1/google-news
- Plans: Starter $29/mo (10,000 results), Pro $99 (50,000), Business $299 (250,000), Enterprise $999 (1,500,000)
- Pay-per-event alternative on the Apify Store: $3 per 1,000 transcripts, $1.50 per 1,000 captures or articles
- MCP: https://mcp.apify.com?tools=siftwright/youtube-transcript-extractor,siftwright/pageframe-screenshots,siftwright/google-news-scraper (Apify token or OAuth)
- Contact: support@siftwright.com

## Docs
- [API docs](https://siftwright.com/docs/): authentication, endpoints, response format, limits, errors
- [Pricing](https://siftwright.com/pricing/): plans, calculator, pay-per-event comparison

## Products
- [YouTube Transcript API](https://siftwright.com/youtube-transcript-api/)
- [Website Screenshot API](https://siftwright.com/screenshot-api/)
- [Google News API](https://siftwright.com/google-news-api/)
- [HTML to PDF API](https://siftwright.com/html-to-pdf-api/)
- [Web data for AI agents](https://siftwright.com/web-data-for-ai-agents/)

## Guides
- [How to get YouTube transcripts with an API (Python, Node.js and curl)](https://siftwright.com/blog/youtube-transcript-api-python-node/): A practical guide to fetching YouTube transcripts from code: why DIY scrapers get IP-blocked on servers, working Python, Node.js and curl examples, languages and translation, SRT/VTT, playlists, and chunking transcripts for LLMs.
- [Feeding clean web data to AI agents with MCP and tool calls](https://siftwright.com/blog/web-data-for-ai-agents-mcp/): How to give AI agents reliable access to YouTube transcripts, web page screenshots and Google News results, through a hosted MCP server or plain tool calling. Setup for Claude, Cursor and VS Code, plus tool design tips that keep context small.
- [Website screenshot API in Python and Node.js: full-page, mobile, retina and PDF](https://siftwright.com/blog/website-screenshot-api-python-node/): How to capture website screenshots from code without running your own headless browser fleet. Working Python and Node.js examples for full-page, mobile, retina and PDF captures, late-loading pages, and a batch script with retries.
- [Build a news monitoring pipeline with a Google News API (Python)](https://siftwright.com/blog/news-monitoring-pipeline-google-news-api/): A step-by-step guide to monitoring brand, competitor and industry news in Python. Design queries, fetch Google News results as JSON, deduplicate, store in SQLite, and send a daily digest, in about 120 lines of code.
- [HTML to PDF at scale: print CSS, fonts, signed URLs and batch rendering](https://siftwright.com/blog/html-to-pdf-at-scale/): A practical guide to generating PDFs from HTML for invoices, reports and certificates. Print CSS that works, page sizes and breaks, waiting for fonts and charts, rendering private documents via signed URLs, and a batch pipeline for thousands of PDFs.
- [Pay-per-result vs subscription API pricing, explained with real numbers](https://siftwright.com/blog/pay-per-result-vs-subscription-api-pricing/): How data API pricing models really compare. Credits, per-request, per-result and subscription pricing explained, the hidden costs to look for, and worked break-even examples using our own two pricing options, including when the cheaper one isn't our subscription.

## Optional
- [Comparisons](https://siftwright.com/compare/)
- [Changelog](https://siftwright.com/changelog/)
- [Custom data feeds](https://siftwright.com/data/)