A YouTube Short can pull in more comments in a day than a long-form video does in a month, and the vertical player shows you almost none of them — a ranked handful, a swipe, and the next Short. If you want to read what the audience actually said, you need the list in a file. This guide covers how to download YouTube Shorts comments free to CSV, Excel or JSON, which links work, why the file is smaller than the comment badge, how Shorts comment sections differ from long-form ones, and what people do with the file once they have it.
Shorts comments are ordinary YouTube comments — with one catch
Under the hood a Short is a normal YouTube video that happens to be vertical and under three minutes. It has the same eleven-character video id, the same comment system and the same moderation rules as any watch page. That is good news for exporting: a youtube.com/shorts/ link, a watch?v= link and a youtu.be link all point at the same id, so any of them can be pasted into the YouTube comment exporter and the result is the same file.
The catch is the player. The Shorts player renders comments in a side panel that is ranked, truncated and lazily loaded, and a large share of the list is never drawn on screen at all. Copy-pasting from the panel gets you a few dozen top-ranked comments and none of the rest. The exporter reads the public comment list directly, so it does not care what the panel decided to show.
Downloading the comments on a Short
- Copy the link. On the Shorts player, tap Share then Copy link. On desktop, the address bar is fine. Tracking parameters and timestamps are ignored.
- Paste it into the exporter. No sign-in, no browser extension and no Data API key. The exporter fetches the public list and shows the count before anything downloads.
- Choose replies or not. Off by default. Shorts threads are reply-light — a typical Short has a handful of replies per hundred top-level comments, against far more on a long-form video — so leave replies off for counting and switch them on for conversation analysis.
- Pick a format. Excel when a person will sort and filter, CSV for import into anything, JSON when a script or a language model reads it next.
The file carries author, username, avatar_url, text, likes, replies, created_at, language, is_pinned, id and reply_to_id, then every extra field YouTube returns for the comment. The same columns come out in all three formats.
Why the badge says 2,300 and the file has 1,900 rows
This is the question we get most about Shorts, and the answer is moderation. The count on a Short includes every comment YouTube accepted at the moment it was posted. Three things then remove comments from the public list without changing the badge:
- Held for review. Channels can hold all comments, or comments that might be inappropriate, for manual approval. Held comments are invisible to the public until approved, and many never are. YouTube’s comment moderation help page describes the settings.
- Blocked words. A channel’s blocked-words list quietly holds any comment that matches. On a Short about a product, a competitor’s name is a common entry on that list.
- Likely spam. YouTube’s own classifier holds link-heavy and repeated comments. Shorts attract far more of these than long-form videos because they are discovered through the feed by accounts that never subscribe.
Deleted comments can also linger in the badge for a while after removal. The result is that the public list — the only list anyone outside the channel can read — is almost always shorter than the number under the Short, and the gap is wider on Shorts than on long-form. If you are counting giveaway entries or measuring response volume, use the row count, not the badge.
How a Shorts comment section reads differently
Once you have a few Shorts exports side by side with long-form exports, the differences are hard to miss, and they change what the file is good for.
Shorter comments. The median Short comment is a few words. A long-form comment section on the same channel runs several times longer per comment. Sentiment and keyword work on Shorts is a counting exercise across many short rows, not a reading exercise.
More emoji-only rows. Because the Shorts player makes it easy to fire off a reaction and swipe, a large share of rows are a single emoji. Sort by text or filter on length and they cluster together. For a giveaway they are not valid entries and the picker removes them automatically; for sentiment they are a signal on their own — a wall of 🔥 says something a long paragraph does not.
Less nesting, more repetition. Few replies, but many near-identical comments, because Shorts viewers do not read the thread before posting. A COUNTIF on the text column surfaces the repeated line — often a question the Short did not answer, which is the most useful single fact a creator can pull out of the file.
Wider language spread. Shorts reach through the feed rather than through subscriptions, so the language column on a Short is usually more mixed than on a long-form video from the same channel. Filtering by language before reading saves a lot of scrolling and, for brands, shows where the Short actually travelled.
What people do with a Shorts comment file
The three jobs we see most, in the order they show up.
Giveaways. “Comment to enter” on a Short is the fastest-filling giveaway format on YouTube. Draw the winner with the free comment picker — paste the same link, set a keyword rule so emoji-only reactions do not count, exclude your own channel, one entry per person — and keep the CSV as the audit trail. The YouTube comment picker guide has the full walkthrough, including which comments YouTube hides from the draw.
Audience and sentiment research. A creator or a brand pulls the comments on their last twenty Shorts, combines the CSVs and counts what recurs: the question that keeps coming back, the product that gets named, the thing people say they wish the Short had shown. The guide to analyzing YouTube comments covers the spreadsheet and AI steps. Pro passes add an AI read of the file, though most Shorts work is done with a sort and a COUNTIF.
Competitor and trend research. Shorts on a trend from ten different channels, exported and stacked, show which framing the audience responded to. Sort each file by likes: the top-liked comments on a competitor’s Short are the closest thing to a free focus group on that competitor.
What we measure on Shorts exports
We operate the exporter, so the facts above come from watching Shorts exports run rather than from reading about them. Three things we have measured. The badge-to-rows gap on Shorts sits above the gap on long-form videos from the same channels, which is why the moderation section above exists; the first support question on nearly every large Shorts export is “where are the missing comments”. The share of emoji-only rows on a Short is several times the share on a long-form video, which is the reason the picker excludes emoji-only comments as entries by default and why we recommend a keyword rule on Shorts giveaways. And the reply ratio on Shorts is low enough that replies are left off by default without losing much; the setting exists for the conversation-analysis cases where it matters.
The column names are the live export columns and the free allowance is the live value. Prices are not quoted here on purpose; the pricing page has the current numbers and there is no subscription. For the platform’s side, YouTube’s comment moderation help explains what channels can hold, the YouTube Data API commentThreads reference documents what the official API returns and its quota limits, and YouTube’s Terms of Service govern how you use what you download.
Getting the file into Sheets, Excel or a script
CSV opens in anything, but two habits keep a Shorts export clean. In Google Sheets use File then Import and set the separator to comma with “convert text to numbers and dates” switched off, so an eleven-digit comment id is not rounded and a comment that is only “1” stays text. In Excel, open the .xlsx download rather than the CSV when comments include non-Latin scripts; Excel guesses CSV encoding badly and Thai, Arabic or Japanese rows can arrive as question marks. For a script or a language model, JSON keeps every field typed and nested, and the id and reply_to_id pair is enough to rebuild threads without a lookup table. The YouTube-to-Excel guide walks through the same steps with screenshots.
Privacy and what not to do with the file
The comments are public; the export is a dataset about identifiable people. Keep it for the analysis you downloaded it for, limit who can open it, and delete it when the job is done. Do not message commenters from the username column, do not publish quotes with names, and if you or your commenters are in the EU, treat the file as personal data under the GDPR with a written purpose. A winner announcement needs a handle and a count, not the list.
Download YouTube Shorts comments free →
Related reading: YouTube comment downloader, export YouTube comments to Google Sheets, YouTube comment scraper.