Gemini 3.5 Live Translate: Can it support a 4-hour meeting and text-only translation output?

Google AI Studio
1

I’m currently testing Gemini 3.5 Live Translate in Google AI Studio for real-time translation during long meetings.

My use case is a meeting that lasts around 4 hours. The speaker will speak in English, and I need Gemini to translate the speech into another target language continuously throughout the whole meeting.

Could you please help clarify the following points?

  1. Is it feasible to use Gemini 3.5 Live Translate continuously for a 4-hour meeting in Google AI Studio?
  2. Are there any session duration limits, token limits, rate limits, or context-window limitations that could interrupt the translation during a long meeting?
  3. I’m using a Google AI Pro account. Does this plan provide higher limits for this Live Translate use case in Google AI Studio, or are the limits still based on the underlying project/model quota?
  4. For my use case, I only need the translated text/transcript to be displayed. I do not need the translated voice/audio output. I already turned off “Echo Target Language”, but the translated voice is still being played back. Could you please confirm whether this is expected behavior, a limitation of the current Google AI Studio UI, or a bug?

This is important for us because we would like to use Live Translate for long internal meetings, but we need to confirm whether it can run reliably for approximately 4 hours and whether audio output can be fully disabled.

Thank you in advance for your support.

2

Hi,

  1. Yes, it should be technically feasible. We recommend testing this over a 4-hour conversation in Google AI Studio to verify. Let us know if you face any issues.
  2. Technical Limits: Model token limits are detailed in the Gemini 3.5 Live Translate documentation. However, Token limits should not interrupt this use case.
  3. Pro Account Limits & Credits: To use higher rate limits, link an API key from a paid GCP project in Google AI Studio. The session will then inherit the rate limits of that project’s billing tier. Note 1: If using Google AI Pro subscription, you can a) claim free subscription credit benefit in the Google Developer Program Dashboard, b) specify a GCP project while claiming, c) import that specific GCP project in AI Studio, d) Create an API key linked to that project and then use that API key for translation. Note 2: For prepay billing accounts, a small initial deposit (such as $1) may be required to activate the above mentioned subscription credits.
  4. Audio Output & “Echo Target Language”: The “Echo Target Language” setting only suppresses audio output when input speech is already in the target language. For example, if your target language is English and input speech is detected in English, the model suppresses audio playback for that portion. It does not mute translated audio output generated from non-target input speech.