I Sat Through a 90-Minute Lecture Letting Gemini Take My Notes — Here's Whether It Actually Worked

There are two genuinely different ways to have Gemini handle a live lecture for you — one listens in real time as it happens, one works from a recording afterward. They're not interchangeable, and picking the wrong one for your situation is the difference between walking out with a usable study guide and walking out with nothing. I tested both. Here's what I found.

Section 1: The Two Real Ways to Do This (and Who Each One Is For)

Every guide that promises "use Gemini to transcribe your lectures" tends to skip the fact that there are two separate features involved, with different costs, different setups, and different results. Picking between them upfront saves you from finding out mid-lecture that the one you picked doesn't fit.

Method 1: Live, real-time capture with "Take notes for me"

Google Meet's Gemini-powered "Take notes for me" feature, originally built for scheduled video calls, now works for in-person settings too — you open the Google Meet app, tap "Take Notes for Me," set your phone or laptop nearby, and Gemini listens and transcribes in real time as the lecture happens, no video call or invited participants required.

The catch: this specific feature requires a paid Google AI Pro or Google AI Ultra subscription on an individual account, or an eligible Google Workspace plan through a school or employer — it isn't available on a free personal Google account. Google also recommends sessions of at least 15 minutes for the AI to have enough context, which a typical lecture comfortably clears.

When you stop the session, Gemini generates a structured Google Doc — discussion points, a summary, and a full transcript — saved automatically to your Drive, with a recap emailed to you shortly after.

Method 2: Record it yourself, then upload to Gemini Notebook (free)

If you don't have a paid plan, or you'd rather not run a live app during the lecture, the free route is simpler than it sounds: record the lecture as an audio file the normal way — your phone's voice memo app works fine — then upload that file to Gemini Notebook afterward as a source. This is the exact workflow covered in our earlier post on analyzing audio files and meeting recordings with Gemini Notebook, and it applies to a lecture recording exactly the same way it applies to a meeting recording.

This route costs nothing beyond a free Google account, handles recordings well beyond the length limits of quick in-app transcription, and gives you the same kind of structured summary, Q&A chat, and even an Audio Overview of the lecture afterward if you want to review it by listening instead of reading — the method from our Audio Overviews for passive learning post.

One important note before you record anything

Check your institution's policy and, where relevant, your instructor's permission before recording a lecture — many schools have specific rules about recording class sessions, and a quick check ahead of time avoids an awkward conversation later. This applies to both methods, since both start with capturing someone else's spoken words.

Section 2: What Actually Happened When I Tested the Live Method

I tested "Take notes for me" during a 90-minute public lecture, phone propped on the desk in front of me, seated roughly two-thirds of the way back in a mid-sized lecture hall — a realistic seat, not a front-row best-case setup.

Where it worked well

The transcript captured the substance of the talk accurately whenever the speaker was reasonably close to a microphone or projecting clearly, which covered most of the session.

The generated Google Doc landed in my Drive within a few minutes of stopping, with a summary section and a full transcript I could search through — genuinely faster than typing my own notes would have been.

Catching up mid-session worked as advertised. I stepped out briefly and came back to check the "summary so far" option, which gave me an accurate recap of what I'd missed without needing to review the full transcript.

Where it fell short

The output format is built for meetings, not lectures, and it shows. This feature is designed around back-and-forth discussion — it tries to organize content into things like decisions and next steps, which maps awkwardly onto a one-directional lecture where nobody's "deciding" anything. The raw transcript and summary were useful, but the structure needed a follow-up prompt to actually feel like lecture notes rather than meeting minutes.

Distance and room acoustics mattered more than I expected. A couple of stretches where the speaker moved away from the mic or the room got noisy produced noticeably shakier transcript sections. Sitting closer, or using a laptop with a better mic than a phone lying flat on a desk, would likely have helped.

It's not a silent, invisible feature. Google is explicit that this isn't meant to run undetected — there's a visible indicator that note-taking is active, which is a reasonable transparency choice, but worth knowing if you were picturing something more discreet than it actually is.

Turning the raw output into an actual study guide

The single most useful thing I did after the fact was ask Gemini directly, in a follow-up prompt, to restructure the transcript into a study-guide format — key concepts, definitions, and a short list of things likely to show up on an exam — rather than leaving it in meeting-note form. That one extra step made a much bigger difference to usefulness than the live capture itself.

Section 3: Step-by-Step — Both Methods

Method 1: Live capture during the lecture (requires Google AI Pro/Ultra or eligible Workspace plan)

1. Confirm eligibility first. Check that your account has Google AI Pro, Google AI Ultra, or an eligible Workspace plan — the option won't appear otherwise.
2. Open the Google Meet app on your phone or laptop before the lecture starts.
3. Tap "Take Notes for Me" from the Meet home screen — no scheduled meeting or invited participants needed.
4. Position your device as close to the speaker as reasonably possible, or at least in a spot with minimal background noise.
5. Let it run for the full session. Use Pause if there's a break, and Stop when the lecture ends.
6. Check your Drive and email a few minutes later for the generated Google Doc with the transcript and summary.
7. Ask Gemini to reformat it. Prompt it to turn the transcript into a study guide, key-concepts list, or flashcard-style summary rather than leaving it as meeting-style notes.

Method 2: Record and upload afterward (free)

1. Record the lecture using any voice memo or recording app on your phone — no special setup needed.
2. Open Gemini Notebook and create a new notebook for the class or topic.
3. Upload the audio file as a source directly from your device.
4. Open the Source Guide for an auto-generated overview once processing finishes.
5. Ask specific questions in chat — "summarize the main arguments," "list any dates or formulas mentioned," "what did the lecturer say about X."
6. Build the notebook over a semester. Add each lecture as a new source in the same notebook so you can eventually ask questions across the whole course, not just one session.

Which one to actually use

• Already paying for Google AI Pro or Ultra, and want notes without recording anything yourself → Method 1.
• On a free account, or want a permanent audio record you control → Method 2.
• Taking a recurring class across a semester → Method 2, so everything accumulates in one searchable notebook.
• One-off talk or guest lecture you just want notes from quickly → either works, but Method 1 needs the paid plan already active before the talk starts.

Whichever route you use, treat the output as a strong first draft rather than a final transcript — verify anything with a specific number, date, name, or formula against your own memory or the lecturer's slides before you rely on it for something graded.

Are you working with a paid Google AI plan already, or would the free record-and-upload route fit your setup better? Let me know in the comments which one you're trying first.