Zoom live translation: private, local captions for any meeting

You can add live subtitles and AI translation to a Zoom meeting or webinar without sending an extra copy of the conversation to a translation service. This guide starts from the beginning, so you do not need previous streaming experience. By the end, hosts and attendees can have translated captions running on a Windows PC, with one or two translated languages shown at once.

Key takeaways

• Hosts can place a caption window over Zoom or share it with everyone.
• Attendees can translate meeting audio on their own PC without asking the host to change any settings.
• The app listens, turns speech into captions, and translates them on the PC.
• OMNI can display two translated languages for each spoken phrase.
• A modern NVIDIA graphics card gives the fastest captions.
• Nothing is uploaded to an extra translation service.
• There are no monthly fees or per-minute translation charges.

What you need

• A 64-bit Windows 10 or Windows 11 PC.
• A modern NVIDIA graphics card for the best real-time performance.
• OMNI Live Translator, with a 30-day free trial that requires no credit card.
• Enough storage to download the languages you want to use.
• The Zoom Workplace desktop app and a meeting invitation.
• Host, co-host, or panelist permission if you want to share captions in a meeting or webinar. The host runs the meeting, while panelists are approved webinar speakers.
• OBS Studio only if your stream or meeting production already uses it.

How to set up Zoom live translation

1. Install OMNI and download your languages. Open the app and go to the Languages & Models page. Select Get more languages, then download the spoken and translated languages you need. This first download needs an internet connection. After setup, OMNI can listen and translate offline on your PC.
2. Choose a speed and quality setting. Start with Balanced. The app will pick a suitable setup for your PC. Choose Fastest if the captions are falling behind during quick conversations. Choose Most accurate if your gaming PC has a capable graphics card and the meeting uses specialist words. You do not need to choose or understand the technology behind these options.
3. Set the spoken and translated languages. Choose the language that people will actually speak. Do not pick a similar language and hope it will work. Then choose target 1, which means the first translated language. Add target 2 if you want a second translation. For an English presentation with customers in China and Mexico, choose English as the spoken language, then Chinese and Spanish as the two outputs.
4. Choose which Zoom audio OMNI should hear. If you are presenting alone, you can let OMNI listen to your microphone. To translate the whole discussion, turn on watch-anywhere mode. This lets the app listen to the sound playing through your PC, including other people in the Zoom call. Attendees can use the same mode. The host does not need to turn on Zoom’s own translated captions.
5. Start a short private test meeting. Zoom lets you start a meeting by yourself, which is an easy way to test without an audience. Speak in normal sentences. Check that the first captions match what you said before judging the translation. Background music, a distant laptop microphone, and several people speaking at once can all cause mistakes. My recurring live-caption pet peeve is blaming the translation when the app heard the original words incorrectly.

6. Choose how the captions should appear. For private attendee use, open OMNI’s movable caption window. Turn on its always-stay-on-top option so Zoom cannot hide it. Then place it below the Zoom video grid. A host can keep this window private or share it through Zoom’s Share button. You can also place the window over a presentation and share the whole screen. This creates a caption overlay — text shown on top of your presentation. In webinars, only hosts, co-hosts, and panelists can share their screens.

   OMNI also provides an OBS browser-source URL. A browser source is a way to show a web page inside an OBS scene, which is the layout used for a stream or recorded production. Paste the URL into an OBS Browser Source if your Zoom production already goes through OBS. Zoom itself does not have a browser-source box. For an ordinary meeting, the movable caption window is much easier.

7. Test the words and speaking speed you expect. Try names, shortened company terms, numbers, and the normal pace of the real call. OMNI waits through a brief mid-sentence pause instead of translating half a thought too early. On a gaming PC with a modern NVIDIA graphics card, the first captions can appear in under one second. The heavier Most accurate option may need a more capable graphics card.

The same setup can be used for Discord voice conversations and YouTube live streams.

Picking languages: English, Chinese, Spanish, French, Arabic, and Japanese

Choose the exact spoken language first. Then use two-language translation when two audience groups need different captions. OMNI shows both translations side by side after each completed phrase.

For English-to-Chinese and English-to-Spanish meetings, the presenter can speak once while both groups follow the same sentence. English-to-French plus Arabic can work well for regional webinars. Arabic reads from right to left, so leave enough width in the caption window for it to display clearly.

Japanese-to-English and Chinese-to-English meetings often work better when the speaker finishes a full sentence before pausing. Important meaning can arrive near the end. Tone matters too. Japanese business speech may be polite, and the English translation should not sound overly casual. Korean captions may also need a respectful tone instead of a direct word-for-word translation. OMNI’s built-in AI translation works with complete sentences so it can use that extra context.

Why run translation locally?

OMNI runs on your PC. It does not upload or log the meeting audio, captions, or translated text. Zoom still carries the original meeting because participants are talking through Zoom. OMNI simply avoids sending another copy of the conversation to a separate caption or translation service.

Local processing also means no monthly translation bill and no charge for every minute you use it. This can matter during a long interview, consultation, class, or quarterly meeting. Once you have downloaded your languages, OMNI can create and translate captions without an internet connection. You will still need the internet to join the Zoom meeting itself.

Running on your PC does not mean replacing words one by one. The app listens to the sentence, turns the speech into captions, and uses its built-in AI translation to create natural text. For legal, medical, or other high-consequence calls, test important terms before the meeting. Use a qualified human interpreter when professional judgment or certified interpretation is required.

How does it compare to Zoom’s familiar options?

Zoom’s built-in captions are the easiest choice when your organization already pays for an eligible plan. Wordly is aimed at managed events with several languages. Human interpreters remain the careful choice when mistakes could have serious consequences.

According to Zoom’s current translated-caption requirements, the feature is included with Workplace Business Plus, Enterprise Essentials, Enterprise Plus, and Enterprise Premier, or assigned through the Translated Captions add-on. Zoom lists 36 languages that can be used as the spoken language. Greek, Norwegian, and Welsh are available only as translated languages.

If your workplace already provides Zoom translated captions, their built-in position and attendee language choices may be all you need. Wordly adds event tools, translated audio, written transcripts, and summaries. Its available hours are reduced as sessions run. Human interpreters need to be booked and usually cost more for each additional language, but they are often worth it when unclear wording could have legal, medical, or financial effects.

FAQ

OMNI works for Zoom hosts and attendees. It does not replace Zoom’s own caption track.

Can a host add OMNI captions directly to Zoom’s caption menu?

No. OMNI shows captions in its own movable window or through an OBS browser source. A host who wants everyone to see the captions can share that window or share a screen with the caption overlay on it.

Can an attendee use Zoom live translation without the host enabling captions?

Yes. Watch-anywhere mode listens to the Zoom audio playing on the attendee’s PC and shows the translation there. The host does not need a paid Zoom translated-caption plan or a special meeting setting.

Which languages can OMNI translate during Zoom meetings?

The app offers more than 30 selectable languages. Open Get more languages to download the ones you need. You can show two selected translated languages at the same time.

Is the meeting audio uploaded for translation?

No. OMNI listens and translates on the Windows PC. It does not upload or log the audio, captions, or translations. Zoom still carries the original meeting according to its own service and your account settings.

How much does OMNI cost after the trial?

The full-featured trial lasts 30 days and requires no credit card. After that, the launch price is a one-time $19.95 lifetime license for one PC at a time. You can reset and move the license to another PC up to three times. There is no subscription and there are no usage fees.

Can OMNI translate a recorded Zoom meeting?

OMNI is made for live audio, but that can include a Zoom recording playing on your PC. If you have a saved recording and need subtitle files to download, use the OmniSubs video subtitle workflow.

Download the OMNI Live Translator free trial when you’re ready to test the setup with your own Zoom audio and language pair.