How to Listen to Gemini Responses
• By Elliott Tong
To listen to Gemini responses, install the Alexandria Chrome extension, then open any conversation on gemini.google.com. Alexandria detects Gemini's model-response custom elements and waits for the spinner animation to disappear (checking CSS opacity, not just attributes). It then injects a play button into the response header controls. Click it to hear that response read aloud with word-by-word highlighting.
Why Listen to Gemini Responses Instead of Reading Them?
Gemini's responses can be substantial, especially when you use Gemini Advanced for research, analysis, or creative tasks. A single response explaining a complex topic or generating a detailed plan can run well over a thousand words.
Listening lets you absorb that content differently:
You can work while Gemini explains
When Gemini is helping with research or explaining a concept, listening lets you take notes, sketch ideas, or read source material while the response plays. Your eyes are free to be elsewhere.
It suits Gemini's formatting
Gemini often structures responses with headers, bullet points, and clear sections. That structure translates well to audio — each section becomes a distinct chunk, easier to follow than a wall of undifferentiated prose.
You catch the full context
Gemini responses often include context, caveats, and alternatives. When scanning, these are easy to skip in favour of the 'main' answer. Listening forces you through everything, which often includes the nuances that matter.
You can compare response versions
When you regenerate a Gemini response or ask for a different approach, listening to both versions back-to-back is faster than visually scanning for differences.
How Alexandria Works with Gemini
Alexandria's Gemini adapter (559 lines of custom code) handles three challenges that generic TTS tools cannot:
Custom element detection
Gemini is built with Angular and uses custom HTML elements like model-response instead of standard divs. Alexandria's adapter specifically targets these elements, finding the message-content inside each model-response to extract the text [GeminiAdapter.js, selectors from siteConfig.js].
Spinner-based streaming detection
Unlike ChatGPT and Claude which use data attributes for streaming state, Gemini shows a spinner animation (avatar_spinner_animation) while generating. Alexandria checks the spinner's CSS opacity and visibility — not just whether the element exists — to determine when streaming is truly complete [GeminiAdapter.js, isStreaming(), lines 338-363].
Native TTS button awareness
Gemini has its own built-in text-to-speech button. Alexandria's adapter looks for the tts-control element and the buttons-container-v2 where it lives, injecting Alexandria's button alongside Gemini's native one rather than replacing it. This gives you both options [GeminiAdapter.js, findAnchor(), lines 315-333].
Multi-response state synchronisation
Each response gets its own play button. When you click one, it shows the playing state while all others show idle. Switching responses mid-playback stops the first and starts the second — no overlapping audio [GeminiAdapter.js, responseButtons Map + _activeResponseElement].
Tips for Listening to Gemini Effectively
Start with research responses
Gemini excels at synthesising information from multiple sources. These research-style responses are often long and structured — ideal for listening while you work on something else.
Use it for long-form content
When Gemini generates articles, reports, or creative writing, listening lets you review the content without staring at the screen. You'll catch awkward phrasing and flow issues that you might miss when reading.
Compare with Gemini's native TTS
Gemini has its own read-aloud feature. Alexandria appears alongside it, so you can try both. Alexandria adds word-by-word highlighting and variable speed control that Gemini's native TTS doesn't offer.
Speed up for familiar content
At 1x, a 500-word response takes about 4 minutes. At 1.5x, it's under 3 minutes. Alexandria's free tier goes up to 5x. Start slower with unfamiliar topics and speed up when Gemini is covering ground you already know.