Open WebUI tools for YouTube transcripts (that work on servers), Google Flights, Hotels, News and Search

Open WebUI's popular YouTube transcript tools fetch captions from your server's IP. YouTube blocks most cloud and VPS addresses, so on a hosted Open WebUI they often fail. The YouTube Transcripts tool in retracn/open-webui-tools fetches the transcript through an Apify Actor that isn't blocked, and returns the title, channel, language and text, with optional timestamps. The same repo has Google Flights, Google Hotels, Google News and Google Search tools. Each is one Python file: paste it under Workspace → Tools, add your Apify token in Valves, and enable it for a model. All five were tested on 6 October 2026.

Open YouTube Transcript Scraper on Apify $1.50 per 1,000 transcripts ($1.20 on Gold and above)

Step by step

  1. Create a free Apify account. Sign up at apify.com and copy your API token from Settings → API & Integrations. The free plan's $5 monthly credit covers about 3,000 transcripts.
  2. Add the tool. In Open WebUI, open Workspace → Tools → +, paste youtube_transcripts.py from github.com/retracn/open-webui-tools and save.
  3. Add your token. Open the tool's Valves (gear icon) and paste the token into APIFY_TOKEN. Set the caption language and the maximum transcript length there too.
  4. Enable it. Turn the tool on for a model under Workspace → Models → Edit → Tools, or per chat with the + button. Models with native function calling work best.
  5. Ask about a video. Paste a YouTube link and ask for a summary, the key points or a quote. The model calls get_youtube_transcript and answers from the transcript.
  6. Add the others. google_flights.py, google_hotels.py, google_news.py and google_search.py install the same way and use the same token.

Do it from code

Tool call the model makes (from Open WebUI)

get_youtube_transcript(video="https://www.youtube.com/watch?v=UF8uR6Z6KLc", include_timestamps=True)

# {"videoId": "UF8uR6Z6KLc", "title": "Steve Jobs' 2005 Stanford Commencement Address", "channel": "Stanford",
#  "duration": "15:04", "status": "ok", "transcriptLanguage": "en", "transcript": "[0:00] ..."}

From an AI assistant (MCP)

Add this MCP server URL to Claude, ChatGPT or Cursor and ask in plain English:

https://mcp.apify.com?tools=automationnation/youtube-transcript-scraper

Alternatives

Other Open WebUI YouTube tools use the youtube-transcript-api library directly. That's free and works from home connections, but YouTube blocks most cloud IPs (RequestBlocked or IpBlocked errors) unless you add residential proxies yourself. Open WebUI can also load YouTube links as documents through its built-in loader, which has the same IP limits. See the side-by-side comparison.

FAQ

Why do YouTube transcript tools fail on my server?

YouTube blocks most requests from cloud and VPS IP addresses, so libraries such as youtube-transcript-api raise RequestBlocked or IpBlocked there. This tool fetches the transcript through Apify instead, so your server's IP doesn't matter.

What does it cost?

$1.50 per 1,000 transcripts on your Apify account; videos without captions are free. Flights are $0.20 per 1,000, hotels $1 per 1,000, news $1 per 1,000 articles and Google searches $5 per 1,000.

Where is my token stored?

In the tool's Valves on your Open WebUI server. It's sent only to Apify.

Can I use Google search as Open WebUI's web search engine instead?

Yes. Choose the External engine in Admin Panel → Settings → Web Search; there's a separate guide for that.

Written by AutomationNation, the developer of YouTube Transcript Scraper. Prices checked October 2026.