Write your screenplay on a board of plot cards.
ScriptScreen is a free, open-source desktop app for writing screenplays. Every scene is a plot card on an infinite board: its place is set by the card's own index attribute; tag it with characters, locations, story details and time of action, then click a tag to filter the board down to a single thread. Scene text is written in a built-in rich-text editor, and can be dictated — speech recognition runs locally on your machine. Export the finished script to PDF.
Linux and Windows. English and Russian interface. Apache 2.0.
🇷🇺 Русская версия: README.ru.md
- Plot cards on an infinite board — zoom and pan; Shift+Click to add a scene.
- Scene order is a card attribute — the card's index field sets its place on the board and in the finished script; to move a scene, change the number.
- Four tag dimensions — Roles, Locations, Details and Action Times, attached to any scene.
- One-click filtering — click a tag chip to show only the scenes carrying it; tags combine.
- Built-in script editor — full rich-text editing of scene bodies, saved automatically.
- Offline dictation — transcribe audio files or live microphone input via whisper.cpp, locally.
- Front matter — title page, logline and synopsis kept with the script.
- Text import — bring in a script written elsewhere: mark the flat text with a handful of tags and it becomes cards, roles, locations, details and action times, descriptions included.
- Object references in the text — the name of a role, location, detail or action time attached to a card marks itself in the scene text: underlined with a dashed line, showing the description of the object under the cursor.
- PDF export — the assembled screenplay with title page and table of contents, scenes in index order; the description of an object is written into the text at its first appearance.
- Your files stay yours — a project is plain JSON plus HTML scene files on your disk; changing the name or version keeps a separate version of the script instead of overwriting the previous one.
- No account, no cloud, no telemetry.
Open the
releases page,
download the archive for your platform, unpack it and run the install script.
Everything is installed under your user account — no administrator rights are
required. In the file names below <version> is the version of the release you
picked.
| Archive | What it is | Install command |
|---|---|---|
scriptscreen-<version>-linux.tar.gz |
Linux, x86-64. Speech is transcribed on the CPU — works on any machine, no video card needed. | ./install.sh |
scriptscreen-<version>-linux-gpu-vulkan.tar.gz |
The same Linux build, but the speech models run on a Vulkan-capable GPU: dictation, live transcription especially, is noticeably faster. | ./install.sh |
scriptscreen-<version>-windows.zip |
Windows 10/11, x86-64. Speech is transcribed on the CPU — works on any machine, no video card needed. | right-click install.ps1 ▸ Run with PowerShell |
scriptscreen-<version>-windows-gpu-vulkan.zip |
The same Windows build with speech models running on a Vulkan-capable GPU — faster dictation. | right-click install.ps1 ▸ Run with PowerShell |
If you are unsure which one to take, take the plain (CPU) archive: it works everywhere, and the GPU one only changes how fast dictation is, not what the app can do.
A GPU build pays off only on a machine with the matching video card and driver. You can check that before downloading an archive — the GPU test ships inside every archive, but it also runs on its own:
| Platform | Script | Run it with |
|---|---|---|
| Linux | check-gpu.sh |
bash check-gpu.sh |
| Windows | check-gpu.ps1 |
powershell -ExecutionPolicy Bypass -File .\check-gpu.ps1 |
Open the link, download the raw file and run it. The script only reads — nothing is installed or changed — and tells you whether a CUDA or a Vulkan build suits this machine. Run it again from the unpacked archive to also test the whisper binaries the archive brought with it.
The installer also sets up the components used for dictation: whisper.cpp speech
models and ffmpeg, placed under ~/whisper.cpp.
macOS is supported by the code but not published as a build yet — build it from
source with dist/make-dist.sh --platform macos (see DISTRIBUTION.md).
Full installation, uninstallation and troubleshooting instructions:
DISTRIBUTION.md (ru). The same guide ships
inside every archive as INSTALL.md / INSTALL.ru.md.
- User Manual (English) (PDF)
- Руководство пользователя (русский) (PDF)
- Distribution guide — building the archives, for maintainers.
Both manuals ship as PDF inside every distribution archive
(USER_MANUAL_EN.pdf, USER_MANUAL.pdf).
- Open the app and create a project (Project panel ▸ menu ▸ New Project).
- Go to the Cards panel and Shift+Click the board to add your first plot card.
- Add characters, locations, details and action times in their own panels, then tag the card with them.
- Select the card and write the scene in the editor — or dictate it.
- Export the finished script to PDF from the Project panel.
ETH: 0xD28545A4a16C0857cD8af84185CcaAa9Bbef675e
BTC: bc1q20nv9jj0zt04e0sa4h6s8xgyur70e48ue2de5r
SOL: 596MbbR8XGhFbCHzhWGgJEmPe42VjT7bSpMbUPBV41TU
TRX: TSKCpMdPswcqRbBEmnq2cseQ6zr9XiwdDX