How to Export YouTube Comments to CSV, Excel, or JSON

Updated

You can export YouTube comments without the YouTube Data API by pasting the video URL into a comment exporter and downloading the result as CSV, Excel, or JSON. ZocialComment, the tool referenced here, states that this works with no signup and no API keys, and that it supports CSV, Excel, and JSON output. The steps below apply to any public YouTube video; the format you pick depends on what you plan to do with the data.

Step-by-Step: Export YouTube Comments from a Video URL

  1. Copy the video URL. Open the YouTube video and copy the full link from the address bar (or use the Share button). A standard youtube.com/watch?v=... URL is what you need.
  2. Paste it into the exporter. On ZocialComment, paste the URL into the input field. The page describes the flow as "paste any URL and download every comment."
  3. Choose your output format. Select CSV, Excel, or JSON before or after the export runs, depending on the tool's interface.
  4. Run the export and download. The tool returns a file containing the comments. According to the site's own numbers, exports can run up to 100,000 comments when used through its AI-assistant integration.
  5. Verify the file. Open it and check that the row count roughly matches the comment count shown on the video. If it's far short, see the troubleshooting section below.

The site also offers a Chrome extension that lets you export "right from the post" after signing in once, so you can skip copying the URL into a separate tab. That path requires a sign-in; the paste-a-URL flow is described as not requiring one.

Which Format Should You Choose?

Format Best for Trade-off
CSV Spreadsheets, reporting tools, quick sorting and filtering Plain text; no nested structure
Excel Client reports, sharing with non-technical teammates Heavier file; less friendly to code
JSON Sentiment analysis, data pipelines, keeping the full API-style payload Not readable in a spreadsheet without conversion

ZocialComment describes two output modes that matter more than the file type: Clean mode for client reports and Raw mode when you need the full API payload for sentiment analysis. A researcher quoted on the site (Jake Morrison, University of Oregon) says the JSON format made a 200+ video bulk export easier to feed into a data pipeline. If your goal is a report, Clean + CSV or Excel is the simpler choice; if you're writing code, Raw + JSON preserves more fields.

Exporting Comments from Many Videos

For bulk work, the practical approach is to export each video separately and concatenate the files, adding a video_id column so you can trace comments back to their source. The site's MCP integration is the other route: connect ZocialComment to Claude or ChatGPT once, then ask it to export comments from a given video and summarize them. That path is stated to work on exports up to 100,000 comments, which is the only scale limit mentioned in the source material. No batch-URL feature is described, so don't assume you can paste a channel or playlist link and get every video at once.

When the Export Fails or Comments Are Missing

  • Comments are disabled on the video. If the creator turned comments off, there is nothing to export. Check the video page first.
  • The video is private, age-restricted, or deleted. Only publicly accessible videos can be scraped.
  • You're seeing fewer comments than YouTube displays. YouTube's counter includes replies; some exporters return top-level comments only. Check whether your file includes reply threads before assuming data is missing.
  • The URL is malformed. Shortened youtu.be links and URLs with extra tracking parameters sometimes fail. Use the canonical youtube.com/watch?v= form.
  • Very large videos time out. If a video has tens of thousands of comments, retry, or use the AI-assistant route that the site says handles up to 100,000.

Using the Exported Data

Once you have the file, CSV and Excel open directly in any spreadsheet for sorting by likes or date, filtering for questions, or counting mentions of a product. JSON is the better input for a script: load it, extract the comment text field, and run it through a sentiment or topic model. If you're feeding a BI tool or a reporting template, export in Clean mode so you're not carrying unused API fields. For a one-off "what are people asking?" pass, the MCP route lets you skip the file entirely and ask the question directly in Claude or ChatGPT.

Pricing and free-tier limits aren't specified in the source material beyond a testimonial mentioning a "$14 pass" and a "free tier" to test with, so check the pricing page before committing to a large job.