← Apps

Comment analysis

YouTube comment analysis for audience insights

Turn public comments into themes, sentiment, and engagement signals. Paste a link, adjust scope and report options below, then run. Fetches up to 1,000 comments per run (your batch setting). Next page continues the same YouTube sort and date window; you can change batch size, likes filter, sections, and guidance between pages.

Compliance: Comment listings use YouTube's public data; insights are generated by YouTility via Google Gemini (not by YouTube). By using this tool you agree to our Terms of Use (including YouTube's Terms) and Privacy Policy.
Analysis setup
ScopeBatch size, sort, dates, and likes affect which comments are fetched.

Applies to each YouTube fetch before filters. You can change batch size between Next page requests.

ReportPresets pick which dashboard blocks run; Custom lists every block.

Open Customize list to pick sections; there, optional chat icons next to each enabled section hold short hints to steer emphasis—evidence stays comments-only (no new facts). Max 100 characters each.

15 sections included (Full report). for per-section hints via chat icons.

Frequently asked questions

What does this YouTube comment analysis tool do?
It loads public comments for a video through this site’s server, applies your filters, and turns the batch into a dashboard. You always get computed summaries such as comment coverage (analyzed vs public total on the video), a comment health scorecard, an analysis coverage funnel, and a per-batch histogram of how many likes each analyzed comment received. You can also choose which narrative sections the model writes—summary, key findings, sentiment, topics, quotes, conversation safety, and more. Hover the section names in the form for a short explanation of each. When a run finishes, you can download the visible dashboard as a PDF.
Is this YouTube comment analysis free, and do I need to configure anything?
The tool is free to use on this site. Paste a video URL or ID, set your options, and run—no account and no YouTube API key to paste in the browser; listing and analysis use the site’s server configuration.
How many comments can I analyze at once, and what is Next page?
Each run fetches up to a batch size you enter as a whole number from 1 through the server limit (often up to about 1,000 comments per fetch, depending on deployment settings). If more comments exist for the same YouTube sort order, use Next page to continue from the listing cursor. For a multi-page chain, comment sort (relevance or time) and the date range from your first run must stay fixed so the cursor stays valid; you can still change batch size, minimum likes, excluding the creator, report sections, and optional Advanced guidance between pages.
Can I filter comments before analysis?
Yes. Set a minimum like count, optional publish date range, exclude the channel creator’s comments, and choose YouTube’s relevance or time ordering for the listing. Only comments that pass those rules are counted as analyzed and sent to the model for the sections you enabled.
What is the difference between comments retrieved and comments analyzed?
Retrieved is how many comment rows were pulled in that batch before your filters. Analyzed is how many rows remained after your inclusion rules and were used for the report, so analyzed can be lower than retrieved. The header coverage figure compares cumulative analyzed comments (across pages you have run in the session) to the public comment total YouTube reports for the video when that total is available.
What are Advanced mode and per-section guidance?
Advanced mode reveals optional short text hints per enabled report section (within a strict character limit). They steer emphasis—topics to stress, tone, or questions—without replacing what the comments actually say. They are echoed back across Next page runs so pagination stays consistent. You can rely on the default behavior and leave them blank.
What does the comment like distribution block show?
It is a fixed bucket chart of how many analyzed comments in the current batch fall into each range of per-comment like counts (for example 0 likes, 1 like, 2–4 likes, up to 25+). It is computed from the same filtered comments as the analysis, not guessed by the model. The companion card shows how much of the batch’s total comment likes sit in the most-liked tenth of comments, plus median and percentile benchmarks—all for that batch only.
YouTube comment analysis for creators — themes, sentiment & audience signals | YouTility