Super

@biljanapesevska Thanks Biljana! Research is mostly what it's for. If you try it on a thread and something feels missing, let me know.

@freedmvtests Thanks! Usernames felt like the wrong thing to collect by default. For research you almost always care about what was said, not who said it.

@abdulgiwa Same reason I built it. The extension does the fetching in your own browser, so Claude spends nothing on that part. Then you paste the Markdown in once. If you use Claude Code or Claude Desktop, there's also a small local MCP server that lets Claude search your exported files and pull in only the comments it needs.

@relmediapro Thanks Remi, appreciate it!

@prospectsignal Per thread. On Free each export is one thread anyway, and you can do up to 30 a day. If a thread has more than 1,000 comments, it stops at 1,000 and tells you how many it left out, so you never get a cut-off file without knowing.

@get100users Thanks! Community research is a big part of why I built it. Pasting a whole thread into Claude and asking what people keep complaining about beats scrolling through it yourself.

@cotletdev Thanks! Yes, just not with indentation. Every comment gets a number like [c12], and each reply says which one it answers, like "reply to [c3]", so Claude can follow the chain at any depth. Indenting broke in a lot of Markdown viewers (deep replies turned into code blocks), so we went with the explicit link. The OP's comments are marked too.

@vikramparikh90 Thanks Vikram! That's the idea. Try it on your own threads first, and only pay if you need bigger threads, several tabs at once or more AI briefs.

@main0I Files you download are yours, so no limit there. Inside the extension, recent exports stay saved in your browser for 30 days: the last 10 on Free, the last 50 on a paid plan. There's no .xlsx export, but the CSV is set up for Excel, so it opens with the columns and any non-English text intact, and you can save it as .xlsx from there.

@selzee Thanks Simon! I'd rather people see the limits before they install than run into them halfway through a thread.

@lihuodev Thanks! Yes, they do. Every format writes comments in thread order, with each reply right after the comment it answers, all the way down. The CSV also has parent_id and depth columns, so the tree is still there after you sort or filter. The nested JSON keeps replies inside their parent, and in Markdown each reply says which comment it answers, like "reply to [c3]".

@wcywl59 Thanks! A short-reply filter for the Markdown export is something we're planning right now. Braam asked for it above too, so it's near the top of the list.

@vishalmishra1029 Thanks Vishal! If you try it on a thread, I'd like to hear how it goes.

@mtougui Thanks Mehdi! The three numbers are there because Reddit doesn't always send every comment it counts (mostly deleted ones), so I'd rather show the gap than call an export complete.

@aka6712303 @saasfilter Really clear, thanks. I'll add a short-reply filter to the Markdown export first, aiming to ship it this week. I'll reply here when it's live.

@saasfilter Thanks Braam! Noted on the short replies. Just so I get it right: do you mean dropping one-liners like "this" or "+1", or anything under a set length? For now, the CSV has the comment text and score in their own columns, so you can filter those out in a sheet. I know Markdown is nicer for notes, though.

@aka6712303 Thanks Akira! Research and competitor digging is exactly what I built it for. If you run it on a thread and something's missing or awkward, tell me and I'll look at it.

@johnidgi52 Thanks! On long threads it depends on the plan. The free brief reads the top 300 comments by score and says so at the top, so you know what it skipped. With credits it reads the whole thread: it splits it into parts of up to 1,000 comments, keeps each reply chain in one part, then merges the results. Quotes are copied from your own export by comment number, not written by the model, so you can check every point against the source.

@editpsd Thanks, appreciate it!

@creative_studio_stars It's not trying to be vidIQ. It just answers one question, whether a channel is monetized, and shows the signals behind it. Free, no login.