You know the moment. The call starts, everyone says hello, and then a tile appears: "Recording bot has joined the meeting." The conversation stiffens. Someone asks what the bot is for, someone else wonders where the recording will end up, and the candid discussion you hoped for becomes a careful one. Many companies now block meeting bots outright, and plenty of clients simply refuse to talk with one in the room. Yet the need that brought the bot in the first place has not gone anywhere: somebody still has to remember what was said, who agreed to what, and what happens next.
There is a quieter way to get a transcript, and it does not involve inviting anything to your meeting. The upcoming RealtimeVoiceKIT browser extension for Chrome and Edge captures the audio of the browser tab where your meeting is running and transcribes it live. Google Meet already plays every participant's voice through that tab. The extension simply listens to what your browser is already playing, right on your own computer. No bot joins the call, no extra name appears in the participant list, and nobody has to approve a third-party guest.
Using it is deliberately boring. You join your Google Meet call as usual, click the RealtimeVoiceKIT icon in your toolbar, and start a tab capture. From that moment the extension turns the meeting audio into text in real time, so you can glance at the live transcript whenever you lose the thread. Because the capture belongs to your browser and not to the meeting, it works without calendar access, without admin approval, and without asking the other side to change anything about how they join.
The live transcript is only half of it. While the session runs, an in-session AI assistant, powered by frontier AI, is available for the questions you would normally scribble in the margins. Ask for a quick summary of the last ten minutes before you speak. Ask for the action items agreed so far. Ask for a suggested reply when a tricky question lands in the chat. Instead of splitting your attention between listening and note-taking, you stay in the conversation and let the transcript carry the memory.
A detail that matters in practice: the session does not depend on the popup staying open. Close the popup, switch tabs to look something up, present a document, come back. The capture keeps running in the background until you stop it from the toolbar. When the meeting ends, the full transcript is already waiting in your RealtimeVoiceKIT library.
That library is where the transcript starts earning its keep. Every capture arrives with an AI summary, so the sixty-minute call becomes a paragraph you can paste into a follow-up email. You can ask questions about the conversation in plain language and get answers grounded in what was actually said. If the call happened in another language, or needs to travel to colleagues who speak one, you can translate the transcript into more than 50 languages. And if you need captions, SRT and VTT export is one click away.
It is worth being clear about the difference between this and a bot, because it changes more than the mood. A bot is a participant: it needs an invite or calendar access, it occupies a seat, it is visible to everyone, and it is subject to whatever meeting policies the host's company enforces. Tab capture is your own browser making a record of what it already plays for you, like taking notes with very good handwriting. One honest note on etiquette and law: recording rules vary by country and by company, so tell people you are transcribing the call and get consent where your local rules require it. A transcript should never be a surprise.
Google Meet is the obvious case, but the same capture works on anything your browser plays. Zoom meetings joined from the browser, webinars, panel discussions, training sessions, a YouTube video you need quotes from: if the tab makes sound, the extension can turn it into text. That makes it less of a meeting tool and more of a general answer to the question "this audio is playing in my browser, why can't I have it as text?".
The extension is not yet live: it is coming soon to the Chrome Web Store, and it will work on both Chrome and Edge from day one. You do not have to wait to get value, though. RealtimeVoiceKIT already transcribes meeting recordings today: upload the file from Meet or any recorder at realtimevoicekit.com and you get accurate speech-to-text with speaker labels, an AI summary, translation, and subtitle export. The free plan includes transcription minutes every month with no credit card required. Create your account now, and we will notify you the moment the extension goes live.
The RealtimeVoiceKIT team writes about audio, AI, and the workflows that turn recordings into reach for the RealtimeVoiceKIT team.