Yes — real-time voice translation can speak a translated version of a Twitch or YouTube stream's spoken audio while you watch, live. That's a different thing from translating the stream's chat sidebar, which is a separate, more common category of tool — worth being clear about which one a given "translate a live stream" tool actually does before assuming it covers what you're after.
Chat translation vs. audio translation — two different problems
Several existing tools that come up for "translate a live stream" work on the chat — reading the text messages viewers post and translating those into your language. That's genuinely useful for following a chat community, but it does nothing for the streamer's own spoken words. Voxis works on the other half: it translates the stream's spoken audio, the actual voice of whoever's talking, and leaves chat untouched. If what you want is to understand what a foreign-language streamer is saying, audio translation is the one that answers that.
How it works for a live stream specifically
The mechanism is the same system-audio capture used for any other source: it reads whatever's playing through your speakers, so it works the same way whether you're watching in a browser tab or a dedicated streaming app. For anyone who only ever watches streams in a browser and doesn't need system-wide capture, the separate Chrome/Edge browser extension covers that narrower case — translating a single tab's audio without touching anything else on the system.
Setting it up
Open a Twitch or YouTube stream as usual, start Video & Game mode, and pick the language you want to hear. There's no stream-specific setup — the same session keeps working if you switch to a different stream or tab.
Handling fast talkers and overlapping voices
Live streams tend to have less predictable pacing than scripted video: streamers talk over chat notifications, react quickly, and sometimes multiple people talk over each other during co-streams. Rapid, overlapping speech can push the translation a little further behind than a single calm speaker would, since short fragments are occasionally batched rather than translated word by word to keep individual translations coherent.
FAQ
Does this also translate the chat? No — this translates the stream's spoken audio specifically. Chat translation is a separate category of tool.
Does the streamer need to do anything on their end? No — capture happens entirely on the viewer's machine; the streamer's setup is unaffected either way.
Is there a mobile version for watching streams on a phone? No — this is Windows desktop software (with a separate browser extension for tab audio). It doesn't currently run on mobile.
Davut Akca, Founder, VoxisLive. More about VoxisLive.