Repository navigation
v0.7.2
Hammurabi is a system tray app that writes live captions for anything your computer plays or hears, with optional translation. Speech recognition and translation run on your machine. No account, no API key, and no audio leaves your computer.
Download
- Windows (x64):
hammurabi-windows-x64-setup.exe - Windows with an NVIDIA RTX 50 series GPU:
hammurabi-windows-cuda-setup.exe(much faster captions and translation; needs a recent NVIDIA driver) - Windows with a GPU (NVIDIA, AMD or Intel):
hammurabi-windows-gpu-setup.exeadds Read aloud in the speaker's own voice (Speaker's voice). Its voice model is for personal, non-commercial use only. - The same Windows builds come as portable
.zipfiles too. - macOS (Apple silicon):
hammurabi-macos-arm64.zip - Linux (x64):
hammurabi-linux-x64.tar.gz
First run
- Hammurabi downloads its speech model on first use (about 300 MB on CPU builds, 1.6 GB for the GPU default
turboon CUDA and Metal). The optional built-in translation model is about 1.7 GB more. - macOS: move
Hammurabi.appto Applications. It is not notarized, so macOS blocks the first launch, and right-click > Open no longer gets past that. Runxattr -dr com.apple.quarantine /Applications/Hammurabi.appin Terminal, or open it once, click Done, then click Open Anyway in System Settings > Privacy & Security. Hammurabi runs in the menu bar. When you start captions, macOS asks to allow the microphone and system audio recording; allow both. If you declined, turn Hammurabi on in System Settings > Privacy & Security > Screen & System Audio Recording, under System Audio Recording Only. When upgrading from an earlier version, remove the old Hammurabi entry there with the minus button first, because the permission does not carry over to a new version. - Windows: SmartScreen may warn that the installer is unrecognized. Click More info, then Run anyway. It installs for your user only, needs no administrator rights, and upgrades in place. Remove it from Settings > Apps.
Full Changelog: v0.7.1...v0.7.2