Sweetly translate
About Sweetly
Built to keep conversations alive
Sweetly Translate began with a personal problem: losing an affordable way to speak naturally with people who mattered to me simply because we did not share the same language.
Why I started
When a call-translation service I depended on was retired in May 2025, important friendships suddenly became much harder to maintain. It felt like standing on opposite sides of soundproof glass: we were still there, but the conversation that connected us was gone.
I knew I could not be the only person facing that problem. Families, friends, partners, customers, and colleagues should not lose meaningful relationships because existing translation tools are difficult to use, limited to captions, tied to one calling platform, or require everyone to install something.
From the first prototype to Sweetly
The first version translated speech in one direction using local speech recognition and speech synthesis. It was free to run, but both people needed the software and computers without powerful graphics hardware struggled to provide accurate recognition and natural voices. It proved that the idea could work, while also showing exactly what had to improve.
After months of research, testing, rebuilding, and following more than a few dead ends, the pieces finally came together: bidirectional spoken translation from one installation, natural voice output, support for the tone and meaning behind what was said, and compatibility with the calling apps people already use.
154 languagesfor spoken translation
One installonly the Sweetly user installs the app
No meeting botkeep using the calling apps you know
Everything you need is included
Sweetly includes its own signed virtual audio drivers. There is no separate virtual-cable purchase, third-party API key, meeting plugin, or extra translation software to configure.
Your call audio stays separate
Dedicated call endpoints isolate the conversation so music, notifications, videos, and unrelated computer audio do not accidentally enter the translation.
Speak naturally
Sweetly is designed to wait until you have finished speaking, helping reduce mid-sentence interruptions and incomplete translations—even with atypical speech patterns.
Works with the apps you already use
Use Sweetly with Zoom, Microsoft Teams, Google Meet, Discord, OBS, and other Windows apps that let you choose audio devices.
Built for voices that are often overlooked
I am a throat cancer survivor, so I understand personally how stressful speech technology can be when a voice is difficult for microphones or conventional systems to recognize. Sweetly is designed to work well with atypical speech patterns and background noise while allowing people to speak naturally.
No speech system is perfect, but every part of Sweetly is focused on making multilingual conversation easier, less stressful, and more human. If it can remove that weight for even one other person, then building it has been worthwhile.
What Sweetly is today
Sweetly is a real-time, bidirectional speech translation app for Windows 10 and Windows 11. It works with Zoom, Microsoft Teams, Google Meet, Discord, OBS, and other apps that allow audio-device selection. You speak your language, the other person hears theirs, and their reply is translated back to you—all from one Sweetly installation.
Private by design
Your call audio is translated live and is not recorded by Sweetly. No conversation data is logged, and no third-party bot joins the call.
Communication should not be restricted by language, culture, location, or the way someone speaks. Sweetly exists to make those conversations possible and to keep improving every time people use it.
Joey Sutherland
Founder and developer