Pre-start check
Before purchasing, the Live tab shows whether Live can start with your current settings and lists anything to fix first, such as the NVIDIA GPU, models, or audio devices.
VoxShift is available on Steam (Early Access) and Microsoft Store, where the listing may appear as Vox Shift. The free Demo is on the same Steam store page, and the Free Trial is on the same Microsoft Store listing. The free version limits Live to 10-minute sessions and English output, but it lets you check PC readiness, audio devices, the setup flow, and latency before purchasing.
Before purchasing, the Live tab shows whether Live can start with your current settings and lists anything to fix first, such as the NVIDIA GPU, models, or audio devices.
Speak a few short sentences and check how long it takes from when you start speaking until you hear the translated voice, in both Fast and Quality modes.
Record one sentence (optional) and check that the translated voice suits your use. Without a recording, VoxShift uses your live voice as the reference.
Verify that the translated voice reaches your voice chat app or game through a virtual audio device such as VB-CABLE.
If you stream, check the subtitle window and the OBS subtitle integration in the Audio tab in your actual scene layout.
After checking your PC, VoxShift picks Standard, Low VRAM, or Ultra-low VRAM. Confirm the balance of speed, quality, and stability in that mode.
VoxShift combines local AI models, your voice, virtual audio routing, and subtitles for OBS on your PC. Start with a short test that mirrors your real voice chat or streaming setup.
The Demo is on the VoxShift Steam store page, and the Free Trial is on the VoxShift Microsoft Store listing. Both are on the same page as the full version.
They have the same limits. Microsoft Store and Steam are separate products, so the full version you buy, and its license, works only with the store you bought it from.
Each Live session stops after 10 minutes (you can start again), translation output is English only, the light model set is fixed, and the soundboard holds up to 3 items. Everything else, including the setup wizard, voice recording, device checks, the Pre-start check, the subtitle window, and the OBS link, can be tried.
Yes. Live translation with reference-voice generation requires an NVIDIA CUDA-capable GPU (Turing or newer, 4 GB VRAM or more); on Radeon, Intel GPU, or no-GPU systems, Live voice generation is unavailable. Speed and voice quality also vary by PC, so test your own PC before relying on it.
Check the Pre-start check result, microphone input, translated voice output, the wait until you hear the translated voice, your recorded voice, and routing to Discord or OBS.