StreamTranslate’s live workflow produces text captions and translations. If you need translated speech, voice cloning or selectable dubbed audio, evaluate a service that explicitly supplies those outputs. Captions can accompany your original voice without replacing it, but they are not an audio-dubbing feature.
Define the audience experience in one sentence
Write a requirement such as viewers read Spanish while hearing the original English, or viewers listen to an interpreted Spanish audio feed. The first needs translated text and a suitable display. The second needs an audio service plus a way to carry that sound to the viewer. A player with a CC control does not automatically offer language-switched audio.
Palabra’s live-stream page describes both broadcaster tools and a developer API with translated speech features. Those are vendor-specific capabilities. Use the distinction to ask better questions about your planned output, not to assume every captioning service exposes an equivalent API, listener widget or cloned voice.
Compare the routing and editorial work
| Format | What changes | What to rehearse |
|---|---|---|
| Translated captions | Readable text accompanies the original audio. | Meaning, placement, language choice and timing. |
| Spoken interpretation or dubbing | A translated voice is delivered. | Audio mixing, turn-taking, voice rights and synchronization. |
| Multiple audience audio feeds | The player or service offers distinct listening outputs. | Player selection, supported languages and stream delivery. |
When requesting a demo, use a real sample from your intended show. Ask to see the exact destination player and controls, not only a provider’s internal preview. A good text result does not verify the separate audio mix.
Build the caption route with StreamTranslate
- Connect your microphone, OBS speech source or supported phone feed.
- Set the language being spoken and the desired caption output.
- Use the Twitch extension when viewers should choose individual caption languages.
- Use a Browser Source when everyone should see your selected text in the video.
- Use the same-event YouTube HTTP feed when you want one selectable CC language.
Your original broadcast voice stays in place. The current OBS plugin forwards audio to StreamTranslate rather than replacing the program soundtrack. Keep music and other sources out of the speech feed where practical, and test the text against the actual destination video.
Budget time for a meaningful acceptance check
Compare a short explanation, a fast exchange, a proper noun and a pause. Ask a reviewer who understands the target language whether the meaning survives, not simply whether translated output appears. Check how much delay the audience can tolerate and measure it in the actual player instead of inheriting a vendor’s headline claim.
If you need dubbing later, keep it as a separate delivery requirement with its own test and permissions. For an already recorded video, decide between subtitle localization and a voice-replacement workflow before paying for processing. StreamTranslate’s recorded-video tool also creates subtitles; it does not add voice cloning or lip synchronization merely because the media is no longer live.
Quick answers
Can translated captions replace an interpreter?
They can help people follow speech, but they do not provide spoken interpretation or every requested accommodation. Choose the format around the audience’s actual needs and the event’s requirements.
Does StreamTranslate have a public speech-dubbing API?
This guide describes the supported caption product, not a public live-dubbing API. Do not design an integration around unverified endpoints or features.
Can I add captions while keeping the original voice?
Yes. That is the purpose of the text-caption workflow. Choose the output destination and verify readability and timing while the original audio continues normally.
Continue with your setup
Sources and further reading
Reviewed October 28, 2026. Follow the linked provider instructions for current availability and setup requirements.