Voxta

Voxta 148 Beta: UI Overhaul & Smarter Asset Management!

Added 2025-06-28 14:23:12 +0000 UTC

Hey everyone,

Beta 148 is here, and it's packed with quality-of-life updates that make Voxta smoother, smarter, and more intuitive to use! This update focuses on streamlining your workflow with a UI overhaul, major improvements to asset management, and a host of fixes that address your feedback.

✨ UI & Experience Overhaul

We've focused heavily on making the Voxta interface more powerful and user-friendly.

Advanced Asset Management: You can now explore assets in collapsible folders and instantly play back video and audio files directly within the assets tab.
Favorites Highlighting: Your favorite characters and scenarios are now highlighted in a different color, making them easier to spot.
Smarter & Cleaner UI: We've improved the scenarios list grid, enhanced the avatar view with narrator portraits and better line breaking, and added sleek new toggle animations for services. Drop-down menus have also been restyled for a cleaner look.
Enhanced Diagnostics: The diagnostics page now includes details on the OpenRouter provider and its costs. You can also get a diagnostics link for any message in the chat, not just the last one.
More Control & Insight: The Speech-to-Text playground now shows the recognition end reason, and you can see post-character notes in the portrait view. For OpenAI Compatible services, you now have the option to disable stop words.

⚙️ Core & Server Enhancements

The server is now more flexible and powerful, with new features for both users and creators.

Easy-Update Settings: You can now use Data/appsettings.User.json to keep your custom settings safe when you update Voxta, making migration a breeze.
New Author Name Field: To avoid doxxing yourself when creating and sharing content, you can now set an author name.
Video Streaming Support: We've added support for asset range requests, which allows for video streaming.
Scripting Upgrades: Scripts can now trigger animations even when there’s no voice audio. We've also added case-insensitive Regex support in matchFiles for more powerful scripting.
Performance & Stability: This update includes a potential performance boost for all named pipes modules (like ExLlama, F5TTS, and Orpheus), updated packages (exllama3 0.0.4, kokoro-onnx 0.4.9), and reduced log noise when launching KoboldAI.

🖥️ Desktop App Polish

The desktop experience has been refined for better stability and usability.

Improved Instance Management: The app now prevents you from opening two instances at once. A new menu allows you to easily toggle the console and enable or disable the "minimize to tray" feature (which is now off by default).
Key UI Fixes: We've fixed the dark mode issue in the title bar, corrected a bug where dropdowns would show selected text, and resolved a COMException (0x8007139F) crash.

🛠️ Key Fixes

Content Creation: Fixed a bug where cloning a character didn't copy their images. We also fixed issues where AI-generated thumbnails for scenarios and characters weren't being saved.
Chat & Transcription: Incomplete sentences in non-narrated stories are now correctly stripped, and transcriptions ending with a comma no longer have an extra period added.
Services: Errors from 11labs are no longer hidden, and deleting API keys and Apps now works as expected. The F5-TTS vocos model now correctly shows as downloaded, and the Orpheus default voice now allows custom values.

Important Notes: 🛠️

We're always looking for your feedback! Let us know how the new UI and asset management features feel.
Hit us up on Discord or comment here with any feedback or issues.

Links: