Msty 1.3

We’re excited to announce the release of Msty AI version 1.3, packed with powerful new features and enhancements designed to boost your productivity, streamline workflows, and provide deeper customization options. Let’s dive into what’s new in this version!

Major New Features

1. Export Chat Messages to Markdown or JSON

Now you can export entire conversations with ease in Markdown or JSON format. Aurum license holders can also cherry-pick specific messages for export, giving you more control over the data you need to keep.

2. Azure Open AI Integration

Aurum users can now integrate Azure Open AI as a remote provider, unlocking access to another powerful AI source, giving you flexibility in how you deploy AI models.

3. Live Documents & YouTube Attachments in Chats

Attaching live documents and YouTube videos directly into chats is now possible, enabling richer conversations with more dynamic references and media.

Chat with YouTube videos by dropping the video URLs in the chat input attachments

4. Choose Search Provider for Real-Time Data

Select between Google, Brave, or Ecosia as your real-time data search provider, giving you the choice of how you interact with and retrieve live data within the app.

5. Advanced Real-Time Data Options

Refine your real-time data searches with advanced options, such as limiting searches by domain, filtering within date ranges, and creating custom search queries—ideal for more specific or precise results.

6. Model Selector Enhancements

  • Pin Models: You can now pin frequently used models for quicker access.
  • Model Templates: Apply custom model templates to Local AI models easily from the model selector for tailored configurations.

7. Improved Chat Interface

We’ve introduced an overflow menu (…) for chat messages, featuring more descriptive option labels, along with new keyboard shortcuts for editing and regenerating messages, applying context shields, and more.

8. Enable or Disable Markdown Per Message

Markdown rendering can now be toggled on or off per message, offering flexibility in formatting for those more technical chats or clean conversations.

9. Save Chat from Vapor Mode

No more worrying about losing important chats when in Vapor Mode—now you can save those conversations for later reference.

10. Capture Local AI Logs

When working with Local AI, you can now capture service logs for easier debugging and insights into model behavior.

Key Improvements

1. Enhanced Multi-Device Sync

Workspace API keys can now be used across multiple devices, simplifying setups for those working in diverse environments.

2. Smarter LaTeX Support

We've improved the prompt system to better handle LaTeX, ensuring mathematical expressions render more cleanly and consistently.

3. Sync Across Splits

Key settings like model instructions, context shields, sticky prompts, selected knowledge stacks, and attachments now sync more seamlessly across split chats, improving your multitasking experience.

4. File Type and Knowledge Stack Enhancements

Support for more file types in Knowledge Stacks has been added. Additionally, users can now compose only new changes in their Knowledge Stacks, meaning that after the initial re-compose following this update, only files modified since the last compose will be processed, making future compositions more efficient.

Compose new changes only or recompose a Knowledge Stack

5. Improved Code Blocks

We’ve added a "copy code" button at the bottom of code blocks for easier use and improved the loading of chats with multiple code blocks.

6. Edit Local AI Service Port Number

You now have the ability to edit the port number for your Local AI service directly from the app settings.


This update is packed with improvements aimed at making your experience smoother, faster, and more tailored to your needs. We can't wait for you to try it out—let us know what you think of the new features and how they improve your workflows!

Click here to read the full changelog.

Haven't downloaded Msty yet?

Get started today. Available for Windows, MacOS, and Linux.