SoulScript — Offline AI Voice for Visual Novels
A downloadable tool
SoulScript
SoulScript is a desktop application built for visual novel creators who want a fast, organized way to generate character voice audio using modern AI text-to-speech models.
Instead of juggling multiple tools, scripts, and manual audio workflows, SoulScript brings voice creation, editing, validation, and export into a single offline pipeline designed specifically for dialogue-driven games.
Everything runs locally on your computer — your scripts, voices, and generated audio stay private.
Designed for Visual Novel Workflows
SoulScript focuses on the real needs of VN developers and mod creators:
• Detect characters and dialogue directly from supported game scripts
• Assign unique voices to each character
• Control emotional performance per line
• Generate voice files in organized batches
• Deploy audio back into your project with minimal setup
Supported engines currently include:
• Ren’Py
• Unity (CSV / dialogue file workflows)
Additional adapters are planned for engines like Ink and Yarn Spinner.
The goal is simple:
Reduce the technical overhead so creators can focus on storytelling instead of toolchains.
Local-First by Design
SoulScript is not a cloud service.
• AI generation happens entirely on your machine
• No subscriptions or usage limits
• Your project files are never uploaded
• Internet access is only required during installation or when downloading runtime components and models
Once installed, your entire voice pipeline runs offline.
Built for Creators — Not Engineers
You don’t need to configure complex pipelines or manage external scripts.
SoulScript provides:
• Voice design using natural language descriptions
• Per-line performance instructions
• Waveform preview and validation tools
• Batch generation with GPU acceleration (CPU supported)
These tools are designed to feel like part of the creative process, not a technical hurdle.
Works Across Game Projects
While SoulScript integrates directly with supported engines, all generated audio files use standard formats, so they can be used anywhere.
Whether you’re:
• building a full visual novel
• adding voices to an existing project
• creating a voice mod
• experimenting with interactive dialogue systems
SoulScript helps you move from script → playable audio faster.
Try It in Demo Mode
SoulScript includes a built-in Demo Mode so you can explore the full interface before purchasing.
Generation is limited to the first 40 lines of a script, allowing you to test the workflow without unlocking the full output.
Learn More
Official site:
https://getsoulscript.app/
Included Resources
The download also includes SoulScript Cheat Sheets to help you get the best voice results:
• Prompt formatting guide
• Phonetic editing reference
• Style injection examples
These guides help you fine-tune pronunciation, delivery, and character performance.
Download
Install instructions
Download and run SoulScriptsetup.exe
OR
Unzip SouScriptSetup.zip and run SoulScriptSetup.exe
Requires 15GB disk space.
Development log
- SoulScript v1.1.0 Release23 hours ago
- SoulScript v1.0.1 update18 days ago
- SoulScript — 4 Days After Launch (A Quick Dev Update)20 days ago
Comments
Log in with itch.io to leave a comment.
What you've done is amazing! And it's offline. I don't know how many voice packs it has, but it would be great if it had a good number of voices. I'm going to try it right now; it looks really good.
Thank you — I really appreciate that! 😊 Yep, everything runs locally by design, so creators keep full control and don’t have to rely on subscriptions or cloud limits.
Right now SoulScript focuses more on designing voices than shipping huge preset packs — the goal is to let you shape a voice that fits your characters instead of picking from a fixed list. I’m continuing to refine the voice workflow though, so feedback from people actually using it helps a ton.
If you run into anything while testing, feel free to share — and thanks again for giving it a try!
The results are decent, but we need Voice Cloning to be able to actually use it. Right now, designing a voice with the text prompts is not really great. With voice cloning, we could use other services to design the voice and then use the output from there to clone our voice in soulscript.
Really appreciate you sharing your thoughts — feedback like this helps a lot 👍
I’ve actually been working on a pretty big update behind the scenes focused on improving the workflow and making things smoother overall. I don’t want to overpromise specifics yet, but the goal is to make SoulScript easier to use and less “fight the tool” while staying true to its design direction.
I’m also putting together a small pack of SoulScript cheatsheets that cover some tips and workflow guidance people have been asking about. Once I finish polishing them, I’ll drop a download link here on this page — hopefully they help bridge the gap while the next update rolls out.
Thanks again for taking the time to try it out and leaving feedback 🙂