Developer posts and news about this game.
A faster, cleaner realtime voice changer The realtime voice changer is now 2x faster for everyone, so your speech comes out even closer to real time. Even faster in mainland China: audio there now generates in half the time. It also sounds less choppy, with fewer strange artifact
Ultra-realistic is now about 3x faster Ultra-realistic voices sound more natural and expressive, and they are now about 3 times faster. The offline voice changer is back It is back, and more stable. It runs on your own graphics card and uses no internet bandwidth. Pick "Realtime
A new custom voice studio Design a voice from a picture: upload a picture of your avatar, or snip one straight from your screen, and the studio creates the voice your avatar would have. The picture becomes the voice's face. Start from your own voice: read a short line and hear yo
Voice Changer: free custom voice training New optimized training mode for the live voice changer: train your custom voice for free, on any plan, with about 95% of the quality. Want maximum quality? The original ultra-realistic training is still available and works the same as bef
Local mode is gone for now. Most people could not actually use it: it needed a very powerful graphics card, took up many gigabytes of space, and brought a lot of bugs with it. Compared to everything else in the app, it was also very rarely used. Keeping it was making the app wors
Fixes Fixed a bug where live captions did not work in some countries. Fixed a bug that could leave the app stuck on the loading screen. Echo is easier to find "Hear it while you talk" (echo) now lives in the Speak tab, so you can turn it on and off in one click. Asia server Added
New The echo toggle now shows up in the Speak tab when you are in Voice Changer mode with the online fast or local engine selected, so you can hear your own converted voice. Training a new custom voice for the live voice changer is now a clearer, more intuitive flow. Fixes Fixed
New Asia server Added a server in Asia, so generations in both Voice Changer and AI Voice modes are much faster if you are in the region. Custom voices You no longer have to wait for the portrait to finish generating before using a new custom voice. You can use it right away. You
Improvements Better voice playback, smoother microphone switching, and quicker recovery when your connection drops. Fixes Fixed issues with uploading and training custom voices. Fixed several crashes, and cases where no audio played at all.
Custom voices, now online Your custom voices now work with the online Voice Changer and with AI Voice (online fast) mode. One thing to know: training a new voice so it works with the fast online voice changer uses credits, about the same as 2 hours of ultra-realistic use. Microph
New option: run the live voice changer online The live voice changer (manual mode, the one that converts your voice while you are still speaking) can now run on a powerful online server instead of on your own PC. That means your games will not lag while you use it, because all th
Fixes Fixed several issues with the local engines, including cases where it failed to start or stopped working mid-session. If it did not work for you before, it is worth trying again.
Real-time voice changing Manual mode can now work like a "real" voice changer: it converts your voice live as you speak, running about 1 to 2 seconds behind you. For now, this works with the 5 female preset voices, and only in English, Russian, and Korean. The rest of the presets
New option: use your own graphics card Much faster responses and much lower cost: if you have a decent NVIDIA graphics card, you can now choose to run a large portion of the voice engine right on your own machine, so less has to go through a server. While you use it, you spend on
Try things without spending credits You can now test your microphone in the Devices section. Check that your words are being picked up correctly without using any credits. You can also type your own text and hear it in different voices, so you can find the one you like before com
This update focused a lot on cutting delay. Fast responses are a big part of sounding human, and they make it far less likely that people think they are talking to an AI. About 50% faster Your voice now comes back around 50% faster than before. Audio builds while you talk The app
Faster generation Made several changes that should dramatically cut the wait between speaking and hearing your voice. Continuous mode fixed Continuous mode was not working properly. It is fixed now. Pause briefly while you are talking, and once you let go, the audio comes back al
New voice engine The app now runs on a new voice engine. It is faster and much cheaper to run, so your credits last about 3 times longer than before. The voices sound almost exactly as real as they did. Bought credits before this update? You get extra credits to make up for it, a
Fixed bug for first time users which caused virtual mic to not install properly
Voice creation Voice creation was broken for the last 2 weeks. My apologies. If you tried to create a voice during that time, free credits have already been added back to your account. First-time experience New users now hear the most realistic voice from the start, instead of th