You’re preparing a regional update or an international team meeting in Google Meet, but not everyone speaks the same language. Sound familiar?
Google Meet introduced AI speech translation in May 2025 during its beta rollout at Google I/O 2025, complementing its existing AI-powered captions. This addition represents both a significant step forward and a direct response to the growing demand for real-time translation in everyday video calls.
But this new feature is still in the initial phase and only covers a pair of languages (English <> Spanish). Therefore, if you’re planning a high-stakes meeting, a multilingual town hall, or a global product launch, this built-in tool alone might not cut it. That’s where professional interpretation services and external language solutions like Interprefy come in — giving you the flexibility, support, and quality you need to connect across languages.
Before we jump ahead of ourselves, let’s explore what’s currently possible — and how you can take live translation in Google Meet even further.
As of 2025, Google Meet is used by more than 300 million individuals every month for video conferencing, which includes both personal and professional contexts. Hosts of multilingual Google Meet sessions currently have access to two core types of AI-powered translation technology:
AI-powered captions and subtitles: Google Meet can automatically generate real-time subtitles of what’s being said in a meeting. These can appear in the same language as the speaker, or — depending on the plan and settings — be automatically translated into a different language. For example, if a speaker presents in English, participants can choose to read captions in Spanish, French, or Japanese. This helps bridge language gaps and makes meetings more accessible for everyone.
Google Meet only supports remote simultaneous interpretation through chrome extensions or professional interpretation services such as Interprefy. However, only a month ago, on May 20, 2025, it introduced AI-powered speech translation, allowing participants to have their spoken words translated into another language in near real time. This feature is available in beta for consumers and is planned to roll out to enterprise users later in 2025.
This new feature uses Google’s Gemini AI to listen to a speaker’s voice, translate their words, and deliver the translation using a synthetic voice that aims to closely mirror the speaker’s tone, rhythm, and pacing. A faint echo of the original speech remains audible, making the conversation feel more natural. There is a slight delay — usually a few seconds — as the translation processes in real time.
AI speech translation is currently available to users subscribed to Google Workspace AI Pro and AI Ultra plans. Only one participant in the meeting needs to have the plan active — the translation works for all other attendees, including those on free Google Meet accounts.
At launch, the feature supports English <> Spanish voice translation. Additional language pairs — including Italian, German, and Portuguese — are expected to roll out soon.
During a call, a user with an eligible plan can open the “Speech translation” tool from the meeting menu. They select the language being spoken and the language to translate into. Once activated, participants hear the translated speech instead of the original audio. The original voice remains slightly audible in the background to maintain context.
The introduction of AI speech translation in mainstream video conferencing, such as Google Meet, reflects a wider industry trend toward supporting multilingual communication as a core part of event planning and delivery. As virtual and hybrid events become more global, there's increasing pressure on organisers to make content accessible to audiences who speak different languages.
AI speech translation represents one way that language technology is evolving to support this need — particularly useful for straightforward content delivery or when timelines or resources are limited. Its inclusion in widely used platforms highlights how real-time language support is becoming more accessible to a broader range of organisers and attendees.
Google Meet’s captioning feature automatically transcribes spoken language during a meeting and displays it as on-screen text in real time. Depending on the user’s plan, captions can either appear in the original spoken language or be instantly translated into a selected target language.
Live captions can be turned on by any meeting participant. When activated, Meet automatically transcribes the spoken language and displays subtitles on screen. This can be especially helpful for participants in loud environments or for those who are hard of hearing.
Participants can enable this feature by clicking on More > Settings > Captions and selecting the spoken language. Captions appear at the bottom of the screen and scroll automatically in real time.
For Google Workspace users on eligible plans — such as Business Plus, Enterprise Standard, Education Plus, and Gemini-enabled editions — Meet offers the ability to translate those captions into other languages. When activated, the spoken language is detected and the subtitles are automatically shown in the selected target language.
For example, a speaker presenting in English can be understood by a participant reading subtitles in German, Japanese, or Portuguese. This all happens in real time using AI-powered translation.
As of June 2025, closed captions support over 80 languages, although some are still in beta and may not be available to all users or may be subject to ongoing improvements. Subtitles support over 4,600 language pairs (translation combinations) for users who have access to Gemini for Google Workspace or other qualifying plans.
To improve readability, Google Meet also allows participants to customise how captions appear. Users can adjust the font size, text colour, and background to meet their accessibility preferences. These styling updates became available globally starting in mid‑2024.
While Google Meet has made impressive progress with AI-powered live translation, there are still some key limitations to keep in mind — especially if you're planning a multilingual event with diverse participants or complex content.
Unlike other platforms such as Zoom or Microsoft Teams, Google Meet does not natively support remote simultaneous interpretation (RSI) channels in the way that specialised conference platforms do, meaning it does not provide separate audio channels for interpreters and listeners. There’s no built-in way to connect human interpreters to a meeting or provide multiple audio channels for participants to select from. This can be a challenge for organisers who require live interpretation from professional interpreters in multiple languages.
For professional simultaneous interpretation — where a human interpreter listens and speaks in real time — many organisations use third-party platforms (like Interprefy) alongside Google Meet. These platforms connect with Google Meet to provide separate audio feeds for interpreters and attendees, but this is not a built-in feature of Google Meet itself.
Related Articles:
Google Meet’s new AI speech translation, launched in May 2025, introduces notable advancements in live language accessibility, but it also comes with several limitations in its current form.
To start, language support is limited. Real-time speech translation is currently only available between English and Spanish. Google has announced plans to expand support to Italian, German, and Portuguese in the coming weeks. Although related features such as translated captions support over 69 languages, the spoken AI translation feature remains far more restricted in terms of language pairs.
Access to the feature is limited to users subscribed to Google Workspace AI Pro and AI Ultra plans. It is not available to free users, and wider rollout to enterprise and standard Workspace customers is expected later in 2025. This paywall may pose challenges for users or organisations without access to premium subscriptions, potentially limiting inclusivity.
As the feature is still in beta, users may encounter occasional issues such as minor bugs or inconsistencies in translation quality. There can also be slight latency between the original speech and the translated output, though Google reports this delay is minimal and generally allows for natural conversation flow.
A key aspect of the AI speech translation is its attempt to replicate the speaker’s tone, pacing, and vocal expression through synthetic voice generation. However, this voice recreation may not always capture more nuanced emotional or cultural elements — particularly in complex discussions.
In summary, while Google Meet’s AI speech translation offers promising capabilities, it remains limited by its narrow language support, subscription requirements, and beta status. Future updates are expected to address some of these constraints as the technology evolves and broader access is introduced.
At this stage, Google Meet does not support recordings that include AI-translated speech. The newly introduced AI speech translation feature allows participants to hear real-time spoken translations during a meeting, but these translated audio streams are not saved in the official meeting recording. As a result, the recording reflects only the original spoken content.
For organisers who require recordings with translated audio, third-party tools are currently necessary. This might involve post-processing the original audio using speech-to-text and text-to-speech technologies to generate translated voice tracks, a capability not natively offered within Google Meet.
When it comes to multilingual support, Google Meet currently offers limited assistance beyond the automated features themselves. There is no dedicated support process to ensure that AI-powered captions or speech translation are properly configured or functioning as expected. If a glitch occurs — such as captions failing to appear or speech translation not activating — organisers must troubleshoot on their own, often without real-time technical support. Additionally, there are no pre-event checks, onboarding assistance, or live event monitoring specifically for multilingual features. For events where language access is critical, this lack of proactive and responsive support can pose risks to the attendee experience and the overall success of the meeting.
Google Meet’s AI-powered captions and speech translation tools are a strong foundation for making meetings more accessible — but for many types of events, they don’t cover everything.
When accuracy, nuance, and reliability matter, organisations often turn to professional interpretation services to fill in the gaps. This is especially true for and not limited to:
Large-scale events with global audiences
Multi-language panel discussions where back-and-forth conversation is essential
Meetings with technical, legal, or compliance-heavy content
Events requiring multilingual recordings or post-event materials
Hybrid setups combining online and in-person participants
A professional interpretation partner can support all phases of your event — from sourcing interpreters and testing the setup, to managing features for live handovers between interpreters and ensuring participants receive the right audio or captions at the right time.
While Google Meet doesn’t offer built-in support for remote simultaneous interpretation, there are flexible ways to extend its multilingual capabilities using external tools — without disrupting the meeting experience.
Many professional language service providers now offer solutions that run in parallel with Google Meet. These setups allow organisers to incorporate both human interpretation and AI speech translation into their meetings through companion interfaces or browser-based tools.
Here’s how it typically works:
The original meeting audio is securely routed to professional interpreters or to an AI speech translation engine.
Human interpreters listen to the meeting and deliver real-time interpretation from a remote console, or AI generates a translated voice using synthetic speech technology with expanded language pair support.
Participants can access the translated audio through a browser tab or mobile app — while continuing to follow the meeting in Google Meet.
Live multilingual captions can also be offered in parallel, supporting users who prefer or need to read over listen.
This setup makes it possible to bring advanced language support into Google Meet without switching platforms or asking attendees to install anything. It’s especially useful for organisers who need to support more language combinations, rare languages, technical content, or high engagement across international audiences.
Whether it’s a leadership call, a training session, or a multilingual hybrid event, these external solutions help bridge the gap — adding scalable voice and captioning support to what Google Meet already offers.
While Google Meet offers useful built-in translation features, organisers often need more flexibility, control, and language coverage — especially for high-profile or multilingual events. Interprefy bridges these gaps by integrating seamlessly with Google Meet to provide professional-grade live translation services.
Real-time remote simultaneous interpretation and AI speech translation in over 6000 language combinations, delivered by professional interpreters or AI speech translation engines. Participants can listen to the translated audio through a companion browser tab, mobile app, or embedded player — without having to leave the meeting.
Live multilingual captions can also be added, powered by AI and enhanced with Interprefy's custom vocabulary feature to improve accuracy for event-specific terminology, names, and acronyms.
Interprefy allows Google Meet to scale well beyond its native capabilities — offering professional-grade live translation without asking users to change how they connect.
Google Meet continues to evolve as a platform for global communication, with recent updates like AI speech translation, expanding what’s possible in multilingual meetings.
For everyday conversations, these built-in features can go a long way. But for professional, high-profile or complex events, they often need to be supplemented.
There’s currently no built-in remote simultaneous interpretation in Google Meet, and AI speech translation — while promising — is still limited in language coverage, platform compatibility, and recording capabilities.
That’s why many organisations rely on external support to deliver truly seamless live translation experiences. By combining Google Meet’s strengths with professional interpretation services, AI speech translation tools, and real-time multilingual captioning, it’s possible to create meetings that are clear, inclusive, compliant, and globally accessible — no matter how many languages are involved.