Chat activity across the entire VoD, with six reaction layers underneath. Zoom into any stretch and the peaks tell you where to look first.
On an eight hour stream, scrubbing for highlights takes hours. The chat already did that work. When something funny, scary or impressive happens, message volume jumps. The graph turns that jump into a shape you can scan in five seconds.
You stop scrubbing and start clicking. Pick a tall peak, watch the few seconds before it, and decide whether to keep it.
Strip away the styling and the graph is one number repeated across the VoD: how many chat messages landed in that stretch. Plotted from the moment the stream started to the moment it ended, that number becomes a shape. Nothing is left out, every part of the stream gets its own bar.
Read it the way you would read any heatmap. Tall bars are hot, flat stretches are cold. A calm ten minutes of gameplay is a flat line. The second chat reacts, a bar shoots up. Scanning the whole thing takes seconds and tells you where the stream got loud without watching a single minute of it.
Twitch shows concurrent viewers under a VoD. That number mostly reflects when people tuned in and dropped off, and it says nothing about what happened on stream. This graph tracks reaction instead of attendance. A stream with a flat viewer count for hours can still show two sharp chat spikes, and those two spikes are the moments worth clipping.
The graph is most useful when filtered. A KEKW-only graph shows where chat laughed. A monkaS-only graph shows where chat got nervous. A POG-only graph shows the highlights.
Voding loads global Twitch emotes, the channel emotes, and 7TV, BTTV and FFZ emotes for the streamer you opened. Pick as many as you like and the graph redraws for just those.
Open a six hour VoD. The graph paints quickly because the chat is already stored with the VoD, so there is nothing to process on first view. Filter to KEKW and OMEGALUL and most of the timeline goes flat. The few peaks that remain are the laughter moments. Click the tallest one. The player jumps to that timestamp and the chat log scrolls to that second. Watch ten seconds.
If it is a clip, drag from a beat before the spike to a beat after it. A bracket appears on the graph and the clip opens in the sidebar. Name it, tag it "funny", move to the next peak. You spend your time judging moments, not hunting for them.
Most VoD editors treat chat as decoration: a single image you cannot search, or nothing at all. Scrubbing the source at 8x still takes an hour for an eight hour stream, and you miss things every time you blink. Voding flips the problem. Chat noticed every moment that mattered when it happened, and that record stays after the stream ends. The graph turns the audience into the search index, the peaks are the candidates, and a click is the seek.
A chart of how many chat messages were sent across a VoD. Tall bars mean chat reacted to something on stream, which is usually a clippable moment.
Yes. The graph zooms and pans for streams of any length, including twelve hour subathons.
Yes. Add words, Twitch emotes, or 7TV, BTTV and FFZ emotes as filters. The graph redraws to show only messages that contain at least one of the selected tokens.
The chat log keeps the original Twitch timestamps, so you can zoom in until the bars are individual seconds. That helps for moments that happen fast, like a clutch play that chat reacts to in two seconds, not twenty.
Still wondering about something? The team hangs out in our Discord and answers there fastest.
Seven days free on the 50 GB plan. No card for Twitch accounts older than 90 days.
We use a strictly necessary sign-in cookie always, and Google Analytics only if you accept. See our Privacy Policy.
