Disclosure: I made the Actor used in this article, so I'm biased. Apify links here are referral links. The same steps work with any tool that gives you captions; I note the free alternative below.
You have a YouTube playlist, say a 40-video course, and you want the text of every video. Maybe it's for a ChatGPT or Claude knowledge base, for notes, or to search across a whole series. Copy-pasting 40 transcripts by hand is a bad afternoon.
Here is how to get all of them in one go, with Python, and without a YouTube Data API key or a Google Cloud project. The result is a CSV and one Markdown file you can drop into an AI tool.
What you'll need
- Python 3.9 or newer
- A free Apify account (new accounts include free monthly platform credit, so the example below costs well under a cent; please check the current credit on Apify's pricing page)
- Your Apify API token, from Settings → API & Integrations
pip install -U apify-client pandas export APIFY_TOKEN="your-token-here" Step 1: Run the Actor on your playlist
YouTube Transcript & Subtitles Scraper takes playlist, channel or video links and returns one row per video. Pricing is $0.004 per transcript delivered ($4 per 1,000). Videos with no captions are not charged.
import os import pandas as pd from apify_client import ApifyClient client = ApifyClient(os.environ["APIFY_TOKEN"]) run = client.actor("logicbase_systems/youtube-transcript").call(run_input={ "playlistUrls": ["https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi"], "maxVideosPerSource": 3, # raise this for the whole playlist "languages": ["en"], "includeSegments": False, # plain text only, smaller output }) I used a short public playlist (3Blue1Brown's neural networks series) and maxVideosPerSource: 3 so you can test cheaply. Change the number, or leave it at the default of 20, once you're happy.
Step 2: Read the results
rows = list(client.dataset(run.default_dataset_id).iterate_items()) ok = [r for r in rows if r["status"] == "ok"] print(f"{len(ok)} transcripts, {len(rows) - len(ok)} videos without a transcript") Every row has status. Videos that have no captions come back as status: "error" with a reason such as NO_CAPTIONS, so one bad video never breaks the whole run. Each good row has the title, channel, publish date, word count, language, and the full text.
When I ran exactly this code, I got 3 transcripts back in a few seconds.
Step 3: Save as CSV and as one file for AI tools
df = pd.DataFrame(ok)[["title", "channelName", "publishDate", "wordCount", "text", "videoUrl"]] df.to_csv("playlist_transcripts.csv", index=False) with open("playlist_transcripts.md", "w", encoding="utf-8") as f: for r in ok: f.write(f"# {r['title']}\n{r['videoUrl']}\n\n{r['text']}\n\n") You now have a spreadsheet-friendly CSV and a single Markdown file with a heading and link per video. Upload the Markdown file to ChatGPT, Claude or NotebookLM, or split it into chunks for a RAG pipeline.
Useful options
| Option | What it does |
|---|---|
languages: ["en", "hi"] | Preferred languages, in order. Human-made captions are used first, automatic ones as a fallback. |
includeSrt: true | Adds the transcript as SRT subtitle text. |
includeSegments: true | Adds timestamped caption lines (start, duration, text). |
channelUrls | Same thing for a whole channel: it takes the newest videos first. |
videoUrls | A plain list of video links, mixed freely with playlists and channels. |
Things to know before you rely on it
- It reads captions that already exist on the video, human-made or YouTube's automatic ones. It does not listen to audio, so videos with captions turned off can't be transcribed.
- Automatic captions can contain mistakes, especially with names and technical terms.
- Check that your use of the text respects YouTube's terms and the video owners' copyright.
The free DIY alternative
If you only have a few videos and don't mind some setup, you can do it yourself with yt-dlp:
yt-dlp --skip-download --write-subs --write-auto-subs --sub-langs "en.*" --convert-subs srt "PLAYLIST_URL" That works well from your own computer. The catch is that YouTube often blocks or rate-limits requests that come from cloud servers, so a script that runs fine locally can fail once you deploy it. Handling that (rotating IPs, retries) is the part a hosted tool takes off your hands.
Next steps
- Run the same code on a channel: replace
playlistUrlswith"channelUrls": ["https://www.youtube.com/@3blue1brown"]. - Schedule it on Apify to pull each new video's transcript automatically.
- Send the output to Google Sheets, Notion or a vector database with Apify's n8n, Make or Zapier integrations.
- Ready-made versions: playlist to transcripts, channel transcripts, subtitles to SRT, plain text for AI.
If something doesn't work for you, open an issue on the Actor page and I'll look at it.

