Scriptify's first official release! Scriptify is a free and open-source writing tool for authors and other creatives. It is powered by OpenAI's Whisper models for Speech-To-Text tasks.
Changelog:
- Created a launch window for model selection and downloads.
- Modified transcription logic to allow for "realtime" audio transcription by recording the audio in chunks and then processing it in the background.
- Introduced windowed modes for all platforms (no console windows).
- Updated UI with animations and new shortcuts.
- A ton of bug fixes, including fixing
ffmpegextraction,openai-whisperhidden imports and files, and file path fixes. - Several export formats have been added, though, there are some issues with
<h2>headings. - Added an
information panelfor writing statistics.
You can contribute to Scriptify to help improve it for everyone, or open an issue on GitHub to submit feature requests, or bugs.
*Minimum requirements:
- RAM: 2GB
- VRAM: 1GB
- Storage space: 2GB
- GPU: Optional (for larger models, more powerful GPUs)
- CPU: 2 Core CPU
*Minimum requirements for running the smallest and fastest Whisper model. Larger models require more storage space, VRAM and processing power in order to run properly.
Important
For Linux/Ubuntu users:
The release binary is split into multiple zip files due to GitHub's size constraints. We've tried to reduce the file size during and after build as much as possible, but we could not get it below 2GB which is the maximum allowed for releases. You can download all the parts of the binary and use 7-Zip (p7zip-full) or unzip (unzip) to unzip the files.
Using 7-Zip:
sudo apt-get install p7zip-full7z x scriptify_ubuntu_v1_0_0_beta.zipNote
This release will not be included in Distributables since release binaries are already distributed here.